Handheld Spectrometer and Neural Network Model for Chemical and Biological Point Detection

The present disclosure relates to a handheld spectrometer apparatus for performing chemical analyses in the field. In particular, the disclosure relates to a portable spectrometer apparatus that is connected to a mobile computing device to characterize chemicals and detect contaminants.

There are many factors contributing to fuel contamination. Hydraulic and Polyalphaolefin (PAO) fluids are two of the most prevalent and costly contaminants. Specifically, hydraulic and PAO fluid contamination requires defueling of the aircraft, a 30-45-day loss of mission-critical assets, and over 80 man-hours per contaminated fuel incident. Failure to detect contamination can result in Class A mishaps, fouling of small moving aircraft engine parts, excessive deposits/gums on aircraft engine components, reduced service life, and increased sustainment costs. Additionally, defuel carts cost $81,000 per unit, and collection and testing systems range from $80 to $450, with lab tests ranging from $300 to $3,200. Thus, within the U.S. Air Force, defueling high performance jets or other assets is significantly more complex and costly. Within the commercial sector, biofuels, or Sustainable Aviation Fuel (SAF) may introduce new and unknown contaminants including microbes and fungi that my impact aircraft performance. Reducing sustainment costs of aircraft and other fuel-powered equipment offers several operational benefits. For example, lower sustainment costs enhance the affordability of aircraft fleets, allowing for broader deployment and extended service life, which are crucial for maintaining air superiority, while also increasing aircraft availability and mission readiness. Yet there are no existing analytical instruments to detect and quantify the level of contamination in a field setting.

It can thereby be appreciated that a capability to detect and quantify contaminants in jet fuel, as well as other samples, would meet a critical need in the industry. Beyond aircraft, such a capability offers advantages in any remote or limited environment (e.g., medical examination rooms, hospitals, in the field, etc. to detect, classify, and quantify microbes, etc.).

There is a need in the art for a system and method that addresses the shortcomings discussed above.

In one aspect, a man-portable apparatus for identification of compounds is disclosed. The apparatus can include: (a) a housing including a first compartment and a second compartment; (b) a light source that directs light into the first compartment; and (c) an optical system including a dichroic beamsplitter and a collection lens, wherein the collection lens is oriented along a first axis, and the dichroic beamsplitter is oriented along a second axis that is at a 45-degree angle relative to the first axis.

In another aspect, a method of detecting contaminants in a chemical sample is disclosed. The method includes a first step of sending, from a chemical analysis application (“app”) installed on a mobile computing device, a control signal to a handheld spectrometer apparatus connected to the mobile computing device that causes the spectrometer apparatus to perform a first test cycle involving a first sample. A second step includes receiving, at the app and from the spectrometer apparatus, first image data captured by a photoreceptor of the spectrometer apparatus, the first image data including spectral data for the first sample. In a third step, the method includes passing the first image data to a deep neural network (DNN) model that is trained to detect and quantify, in spectral data, one or more contaminants of a plurality of potential contaminants. In different embodiments, these potential contaminants can include Polyalphaolefin (PAO), sulfur compound(s), synthetic fuel additive(s), hydraulic fluid(s), and microbial compound(s). A fourth step includes determining, via the DNN model and based on the first image data, the first sample includes a first contaminant, and a fifth step includes presenting, via a graphical user interface (GUI) for the app, a notification indicating the first sample includes the first contaminant.

In yet another aspect, embodiments include a system for detecting contaminants in a chemical sample that includes a processor and machine-readable media including instructions which, when executed by the processor, cause the processor to: (1) send, from a chemical analysis application (“app”) installed on a mobile computing device, a control signal to a handheld spectrometer apparatus connected to the mobile computing device that causes the spectrometer apparatus to perform a first test cycle involving a first sample; (2) receive, at the app and from the spectrometer apparatus, first image data captured by a photoreceptor of the spectrometer apparatus, the first image data including spectral data for the first sample; (3) pass the first image data to a deep neural network (DNN) model that is trained to detect and quantify, in spectral data, one or more contaminants of a plurality of potential contaminants that include Polyalphaolefin (PAO), sulfur compound(s), synthetic fuel additive(s), hydraulic fluid(s), and microbial compound(s); (4) determine, via the DNN model and based on the first image data, the first sample includes a first contaminant; and (5) present, via a graphical user interface (GUI) for the app, a notification indicating the first sample includes the first contaminant.

Other systems, methods, features, and advantages of the embodiments will be, or will become, apparent to one of ordinary skill in the art upon examination of the following figures and detailed description. It is intended that all such additional systems, methods, features, and advantages be included within this description and this summary, be within the scope of the embodiments, and be protected by the following claims.

In different embodiments, the apparatus described herein provides a spectrometer instrument capable of detection and quantification of chemical contaminants in test samples while in remote, limited environments (e.g., in situ, on site, at the location, at a bar, in a medical examination room, in a hospital, etc.). The instrument is a low-cost and robust yet highly sensitive device that is configured to function as a smartphone-based accessory (“dongle”). An Artificial Intelligence (AI) neural network can receive and analyze the data captured by the instrument and generate, in the field, results, simulations, and predictions pertaining to the sample. These tests can be performed in the field in mere minutes.

As will be described in greater detail below, this durable, low-cost, man-portable (can be readily carried by a single person without assistance) and handheld (designed to be operated while held in a human hand) spectrometer apparatus for use with mobile computing devices offers a significant shift forward in monitoring and protecting assets from the effects of fuel or other contamination. In addition, the spectrometer apparatus can be understood to be “field portable”. For purposes of this application, field portable refers to a handheld device or instrument that is designed to be easily transported and used in outdoor or on-site environments, like a field, factory, airplane hangar, etc., allowing for analysis of samples directly where they are collected, rather than needing to be taken back to a laboratory. The proposed devices and systems are compact, battery-powered, and rugged enough to withstand harsh conditions.

Furthermore, in different embodiments, the proposed systems and methods incorporate processing and analysis based on a trained Deep Neural Network (DNN) to classify different levels of fuel or other types of contamination with high accuracy, often at a rate greater than 0.125%, as well as a high degree of sensitivity and specificity of microbial detection. Furthermore, as will be discussed below, the system employs a powerful Federated Learning (FL) platform and specifically addresses and overcomes challenges such as user data privacy, management of deep neural network (DNN) model updates, and performing calibrations (e.g., using a Mercury-Argon laser) due to the strong, isolated peaks in spectra within the mobile apparatus.

Spectroscopy is an analytical technique that studies the interaction between electromagnetic radiation and matter. The interactions give rise to electronic excitations, molecular vibrations or nuclear spin orientations, which can then be analyzed with different spectroscopic instrumentation. There are numerous spectroscopy techniques available, including mass, infrared, near infrared (NIR), Raman, gas chromatography (GC), Fourier-Transform Infrared (FTIR), nuclear magnetic resonance (NMR), ultraviolet (UV), and visible spectroscopy.

Traditional GC remains a gold standard in laboratory settings due to its precision in separating complex hydrocarbon mixtures. However, its field deployment is impractical due to the size of the equipment, extended sample preparation times, and the need for controlled environmental conditions. Conventional GC instruments, for example, offer high sensitivity, but are unsuitable for real-time, on-site applications where rapid results are necessary. Furthermore, conventional liquid chromatography methods such as Gel Permeation Chromatography (GPC) have been explored, but their reliance on UV detectors and requirement for consumable solvents make them less feasible for real-world logistics. Other portable solutions have shown promise for field analysis, but still face challenges related to accuracy, sensitivity, and operational complexity when applied to dynamic fuel testing scenarios.

In addition, NIR spectroscopy, which has been investigated for fuel property analysis, struggles with overlapping absorption bands, making it difficult to distinguish specific fuel components. Similarly, FTIR spectroscopy experiences interference from atmospheric water vapor and carbon dioxide, significantly limiting its effectiveness in outdoor field settings. Traditional distillation-based methods, while reliable in laboratory applications, are unsuitable for rapid field deployment due to their time-intensive processes and large equipment footprint. Furthermore, mass spectrometry relies on a large and very expensive instrument that is not suitable for field use.

As such, Raman spectroscopy is usually preferred for examination of subjects in the field. In general, Raman spectroscopy and infrared (IR) spectroscopy are both techniques that allow for the investigation of bonds within molecules, or vibrational spectroscopy. These techniques provide information not only about molecules through the identification of functional groups and spectral analysis of so-called “fingerprints”, but also allow for the qualitative and quantitative analyses of chemical substances in the sample.

While IR works with the infrared region of the electromagnetic spectrum, measuring how much light is absorbed by the bonds of a vibrating molecule, Raman spectroscopy works by the detection of inelastic scattering, also known as Raman scattering, of monochromatic light from a laser, usually in visible, near infrared or near ultraviolet range. To make a transition Raman active, the polarizability of the molecule during the vibration and the electron cloud of the molecule must change positionally. For an IR detectable transition, the molecule must have a dipole moment change during vibration. The spectra also differ, with IR showing irregular absorbance lines and Raman showing a scattered Rayleigh line and the Stoke/anti-Stoke lines.

In general, with IR, the detector compares the frequencies of light entering the sample with the frequencies of light leaving the sample. The ‘missing’ frequencies correspond to energy absorption by the molecule (or by bonds in the molecule), allowing for identification of the presence of specific chemical entities. On the other hand, Raman spectroscopy typically uses a laser in the near-IR (NIR) or visible region of the electromagnetic spectrum that emits a specific single wavelength of light. Rather than being based on radiation absorbance, Raman spectroscopy is based upon radiation scattering (in this case, light). When light hits the molecule, or a bond in the molecule, it is scattered, which changes the light's energy. The difference in the energy of the light entering the sample and the energy of the light leaving the sample can be measured, giving a ‘Raman shift’. The Raman shift depends upon the frequency of natural vibrations within the bond or molecule, with the frequency of vibration affected by the mass of a molecule or atom (when considering a bond).

The interaction of light with a bond can also be affected by the polarizability of the electron cloud. If a bond in a molecule has a large, diffuse, electron cloud it will scatter radiation more readily, giving a stronger signal. The scattered light is then collected at a detector, and the frequency difference between the incoming light and scattered light determined.

When a transparent (or substantially transparent) liquid, gas, or crystal sample is illuminated with a beam of monochromatic light (e.g., wavenumber zero), most of the incident light is transmitted without change, while a small portion of it is scattered within the whole solid angle. Spectral analysis of the scattered light shows that, in addition to scattering without change of wavenumber of the incident light (“Rayleigh scattering”), it further contains discrete components of altered wavenumber.

In general, there are pairs of new lines appearing in the spectrum at wavelengths positioned symmetrically with respect to the excitation wavelength, where the wavenumbers correspond to transitions between rotational or vibrational energy levels of molecular systems. The appearance of altered frequencies (wavenumbers) in scattered light is called the Raman effect or Raman scattering. While Raman scattering in itself is a relatively weak effect, it is always accompanied with Rayleigh scattering with an intensity usually 3-5 orders of magnitude greater. The new components appearing in the spectrum of the scattered radiation at shifted wavelengths are termed Raman lines or Raman bands, and collectively they are referred to as the Raman spectrum. The Raman bands at wavelengths less than the exciting wavelength are referred to as anti-Stokes lines, whereas those appearing at higher wavelengths as Stokes lines.

In different embodiments, the proposed system can be employed to perform Raman Spectroscopy and GC molecular identification testing in the field, allowing for non-destructive, rapid spectral analysis of fuel or other bio-chemical samples. The proposed instrument's ability to detect molecular vibrations will enable the identification of chemical contaminants, such as but not limited to sulfur compounds, synthetic fuel additives, and water contamination, with high specificity. Raman spectroscopy's speed and minimal sample preparation requirements make it ideal for field-based fuel screening, offering real-time qualitative assessments before further in-depth analysis. To further enhance sensitivity, the system incorporates specialized microfluidic chips configured to amplify weak Raman signals, enabling the detection of trace-level contaminants.

In some embodiments, complementing Raman Spectroscopy, GC testing can serve as a quantitative validation mechanism, allowing for hydrocarbon separation and precise fuel composition analysis. A free induction decay (FID) can also be integrated to provide sensitive and accurate quantification of volatile organic compounds and fuel stability indicators. While GC traditionally requires extended processing times, optimizations in sample preparation and injection efficiency will be implemented to enable faster turnaround times suitable for field deployment.

In different embodiments, to enhance the integration of Raman and GC data, Continuous Wavelet Transform (CWT) signal processing can be employed to refine spectral and chromatographic outputs. CWT can perform a critical role in reducing environmental noise, improving spectral resolution, and detecting transient anomalies within fuel samples. By fusing Raman spectral data with GC chromatograms, CWT can create a multi-dimensional analytical profile, significantly improving the system's capability to detect fuel degradation, contamination events, and deviations from performance specifications.

Furthermore, in different embodiments, the sensor fusion framework described herein can be powered by AI-driven ML models, and trained on combined Raman-GC-CWT datasets to enhance predictive accuracy. These models can be used to correlate spectral and chromatographic features with critical fuel properties, such as but not limited to cetane index, sulfur content, viscosity, and flash point, ensuring a comprehensive and data-driven approach to fuel quality assessment. The fusion of Raman and GC data will enable the system to detect anomalies that might otherwise go unnoticed if a single analytical method were used in isolation.

1 7 FIGS.- 1 FIG. 100 170 100 100 110 Details regarding an embodiment of a mobile spectrometer system are now discussed, with reference to. Referring to, an embodiment of a spectrometer kit of parts (“kit”)is illustrated, disposed in a man-portable storage container (“container”)for easy transport. The kitcan include one or more hardware components and/or devices, as well as software algorithms to provide a unified field portable bio-chemical (e.g., fuel and propellant) real-time analysis system. In this example, the kitincludes a first spectrometer apparatus (“first spectrometer”)with a substantially compact and ruggedized form factor optimized for field deployment. In one example, the hardware is configured to meet or exceed MIL-STD-810 standards, ensuring functional integrity under extreme environmental conditions, such as high vibration, temperature shifts, and humidity. This durability will support long-term field applications while minimizing maintenance requirements. Furthermore, its compact and ergonomic design will facilitate deployment in diverse operational settings without compromising analytical performance.

100 130 140 150 160 110 112 112 130 140 112 114 116 114 116 110 1 FIG. In different embodiments, the kitfurther includes one or more sample bottles, cuvettes, or vials (“vials”), one or more microfluidic chip slides (“slides”), and optionally a sample syringeand a miniature GC moduleincorporating micro-FID (e.g., to measure hydrocarbon profiles). In some embodiments, the first spectrometercan include a sample insertion receptacle (“slot”). In some embodiments, the receptaclecan be sized and dimensioned to receive a sample held in one of the vials, and/or can be sized and dimensioned to receive a sample based on one of the slides. In the examples illustrated in the drawings, the receptacleis configured with two adjacent receiving chambers in fluid communication with one another, including a first test chambersized and configured to receive a substantially cylindrical vial, and a second test chambersized and configured as a slot to receive a substantially elongated and planar slide. Thus, the opening for the first test chambercan be substantially round and include a cylindrical three-dimensional shape extending through a depth of the holder, while the opening for the second test chambercan be substantially slit-shaped, each opening supporting a snug, secure fit for their respective sample holders. In other embodiments, the first spectrometercan include two separate receptacles configured to receive each type of sample holder, rather than one receptacle with two merged chambers depicted in.

100 120 110 120 122 110 122 In some embodiments, the kitcan further include a mobile computing device (“mobile device”)such as a standard smartphone that can be connected to the first spectrometer(e.g., via a USB or micro-USB cable). In other words, the spectrometers described herein can be configured as a phone accessory or phone-compatible dongle. In some embodiments, the mobile deviceincludes a spectral device manager and analysis app (“app”)configured to communicate with the first spectrometerand perform sensor fusion analysis and feature extraction to detect (for example) contaminants such as sulfur compounds, and/or synthetic fuel additives, as well as instances where water is present by its effect on another chemical in the sample. In some embodiments, the app's chemometric algorithms can then correlate these features with key fuel properties such as cetane index, sulfur content, viscosity, and flash point, for example to ensure compliance with DoD and ASTM fuel standards. In different embodiments, the appcan provide a user interface (UI) that can present user-friendly features generated by an onboard AI-powered recommendation engine, guiding operators through the analysis workflow while highlighting potential contaminants and/or deviations from regulatory standards.

2 FIG. 100 170 120 110 232 130 242 244 140 110 250 212 212 214 120 Moving to, some of the parts of kitare shown outside of or removed from container, including mobile device, first spectrometer, a first vialof vials, and a first slideand a second slideof slides. At this time, the first spectrometeris in its disengaged configuration, not yet connected to a computing device. It can be observed that the spectrometer includes a forward-facing peripheral side along with a protruding (e.g., male prong/pin) connector element(s). The connector element(s)can be configured to slide into or otherwise be received snugly by a (e.g., female receptacle/socket) corresponding connector portprovided on the mobile device(e.g., a standard device data / power transfer port).

3 FIG. 3 FIG. 110 350 232 338 334 332 334 336 338 As shown in, in different embodiments, the first spectrometercan be connected to or mounted on a mobile computing device such as a smartphone (e.g., Apple iPhone® or iPad®, phones or tablets employing Android®, etc.), in its engaged configuration, which serves to convert or augment the smartphone's capabilities with the tools needed to perform spectroscopy. Once the two components are attached to or connected to one another, a mobile spectroscopy scanner system (“scanner”) becomes available for use. In, the pre-prepared first vialcan be understood to contain a first samplefor analysis, held within an interior translucent or transparent body portion(e.g., or other material through which light can pass), that can be optionally secured or sealed closed with a cap, stopper, or lid portionon its open end. For purposes of reference, the body portioncan be seen to have a first length. First samplecan comprise of a solid, liquid, gas, or any other substance or chemical compound.

4 FIG. 3 FIG. 334 114 114 336 334 110 114 334 334 110 122 110 122 232 114 In, the body portionhas been inserted or deposited into the first test chamber. In different embodiments, the depth of the first test chambercan be sized to correspond to or extend a distance that is at least half of the first length(see) of the body portionin which the sample is stored in order to ensure the sample is fully visible to the scanning equipment within the first spectrometer. Similarly, in different embodiments, a diameter of the first test chambercan be sized and dimensioned to snugly receive the body portionin order to improve stability of the scan and better protect the vial and its contents. Once the body portionis slid into the chamber, in some embodiments, the first spectrometercan be configured to detect its presence and automatically trigger the pre-installed appto begin to collect data via the first spectrometer. In other embodiments, the user can manually trigger data collection via the appat this time. Upon completion of the scan(s), the first vialcan be removed from the first test chamberand set aside.

In different embodiments, the sample may be alternatively presented for testing via other substrates. For example, the proposed embodiments can also incorporate and are configured to work with Surface Enhanced Raman Scattering (SERS) Substrates and Microfluidic Chips. SERS is a technique that can enhance the Raman scattering of a sample by orders of magnitude, enabling detection of trace molecules, by using a metal surface as a sample substrate. In some embodiments, the SERS substrate can be made from a textured fused quartz substrate with gold deposited on the surface, creating plasmonic nanostructures that enable sensitivity down to parts per billion. Because Raman scattering itself is a rather rare event and occurs for only about 1 in 1 million exciting photons, it is often necessary to excite the sample with a high-power light source. Thus, the SERS technique enhances Raman scattering of a sample by, for example, using a metal surface or other SERS materials as a sample substrate.

There are two primary mechanisms that cause this enhancement. The first is an electric field enhancement, which is created by the excitation of localized surface plasmons. When molecules under interrogation are within the presence of this electric field, the probability of Raman scattering increases greatly. The second mechanism is a chemical enhancement due to inter-and intra-molecular charge transfer between the metal and molecule of interest. This makes SERS a non-destructive, highly sensitive technique that can be used to measure extremely low sample concentrations with significantly lower incident laser power in comparison to a standard Raman measurement.

5 6 7 7 FIGS.,, andA andB 5 FIG. 6 FIG. 242 110 116 242 680 One example of this alternate mechanism by which samples can be received and scanned by the spectrometer involving a microfluidic chip is depicted in. In, the first slideis shown above the first spectrometeras it is about to be inserted into the slit corresponding to second test chamber. For purposes of clarity, an embodiment of the first slide, carrying a sample, is illustrated in a magnified viewin, better showcasing its microfluidic components. As a general matter, microfluidics refers to a system that manipulates a small amount of fluid using small channels with sizes between ten to hundreds of micrometers. It involves a multidisciplinary field that relies on molecular analysis, molecular biology, and microelectronics. Microfluidic chips are therefore often referred to as “lab-on-a-chip” devices, and can serve as miniature platforms that manipulate and analyze small volumes of fluids. These chips, which feature molded or patterned micro-channels, integrate various functions, such as mixing, pumping, and sensing, onto a compact substrate, enabling precise control over minute amounts of liquids.

6 FIG. 242 600 602 604 600 600 In, the first slideincludes a substantially planar base portion (“substrate”)with a first sideand an opposite-facing second side. In some embodiments, the base portioncan include a substantially rectangular outer shape, though in other embodiments the shape can differ. In some embodiments, a lower periphery of the base portioncan include cuts or tapered edges to facilitate insertion into the slit for testing.

242 620 620 620 In addition, the first slideincludes an optical windowwhich is configured to facilitate efficient Raman signal acquisition while maintaining the chip's structural integrity. The optical windowembedded into the microfluidic chip is configured to enhance the Raman signal for clarity in trace chemical detection. In different embodiments, the optical windowcomprises materials such as quartz, which are ideal due to their high optical transparency and low Raman background interference. Quartz microfluidic chips have been successfully utilized for Raman spectroscopy, offering clear optical paths and minimal signal distortion. In other embodiments, materials can include glass, silicon or polymer such as PolyDimethylSiloxane (PDS), etc.

620 600 620 610 620 In different embodiments, the optical windowcan be disposed or arranged in a central region of a lower half of the base portionto ensure that the optical windowaligns precisely with the spectrometer's focal point when the microfluidic slide is inserted into the slot to the slot's full depth, allowing for accurate and reproducible measurements. In some embodiments, one or more filter(s)with micrometer-sized holes are often used as part of the optical windowas a passive means by which to sort particles or cells based on sizes of the holes.

630 In different embodiments, the proposed microfluidic slides leverage microfluidic systems as a low-cost, consumable alternative to Surface Enhanced Raman Scattering (SERS) substrates. In some embodiments, these microfluidic slides can incorporate reaction chambers or reaction wellscontaining colorimetric compounds, thereby enhancing the system's ability to detect components like water content and trace impurities.

640 630 640 630 640 640 In different embodiments, the microfluidic slide can include inlet and/or outlet ports (“sample port”)that are in fluid communication with the reaction wellsthat allow the network to connect to the external environment. Liquids or gases can be injected, managed, and removed from the microfluidic chip through passive or active methods. For example, some of these techniques can involve pressure/flow controllers, syringe pumps, or peristaltic pumps. The microscale fluidic chip's channels may have varying inner diameters, typically ranging from 5 to 500 μm, with modern fabrication techniques enabling formation of structures with sub-micrometer precision. The channel network is specifically designed for the desired application and analysis (cell culture, organ-on-a-chip, DNA analysis, lab-on-a-chip, microfluidic droplets, spectroscopy, etc.). In some embodiments, the sample portis a small opening or inlet designed specifically to introduce a liquid sample into the chip's microchannels, allowing for precise manipulation and analysis of tiny volumes of fluid within the device. In some embodiments, the reaction wellsextend down toward the sample portas individual channels before merging and connecting (fluid communication) to the sample port.

630 Furthermore, in some embodiments, the proposed microfluidic slides can also be used to preprocess samples to improve Raman signal clarity, removing noise and stabilizing the analytical environment. The reaction wellscan further complement Raman spectroscopy by enabling additional chemical interactions that provide unique optical or spectral changes for advanced detection.

7 FIG.A 7 FIG.B 7 FIG.B 242 116 110 122 110 122 242 114 Moving to, the first slideis shown as it enters the opening associated with the slitted second test chamber, and fully inserted or secured in. In, the optical window is now fully immersed or disposed within the testing chamber, and no longer visible to the viewer. At this time, the first spectrometercan be configured to detect its presence and automatically trigger the pre-installed appto begin to collect data via the first spectrometer. In other embodiments, the user can manually trigger data collection via the appat this time. Upon completion of the scan(s), the first slidecan be removed from the second test chamberand set aside.

8 8 9 10 FIGS.A,B,, and 8 8 FIGS.A andB 8 FIG.A 8 FIG.B 800 800 802 852 854 852 802 810 800 820 810 800 800 880 890 810 In different embodiments, embodiments of the proposed spectrometer can include an optimized internal geometric configuration and components assembly to streamline the device and enhance its performance. For purposes of illustration, one embodiment of this optimized configuration is presented in. In, an exposed view, without an exterior casing, of a second spectrometer apparatus (“second spectrometer”)is depicted by way of offering the reader an introduction to the instrument's internal components and arrangement. For example, the second spectrometerincludes a housing unit, which in this example has a substantially rectangular prism shape extending between a first endand a second end. At the first endof the housing unit, a collection lensis mounted, which has a first side that faces outward or away from the interior of the second spectrometer, and a second side that faces inward into a first compartmentof the second spectrometer. In different embodiments, the proposed collection lensis selected to achieve peak performance while maintaining a minimal footprint, ensuring optimal focusing and light collection crucial for high-quality measurements.shows the second spectrometerin isolation/inactive mode, andshows the second spectrometerin scanning mode, as a vialcontaining a sampleis positioned directly next to the collection lensand data collection is begun.

820 828 860 852 810 9 FIG. In this example, the first compartmentcan be seen to include a lens setup/optical configuration where a beamsplitter panelis on which a dichroic beamsplitter (filter) is mounted (e.g., see) is oriented at 45 degrees relative to both a first sidewalland the first endin which the collection lensis installed. In different embodiments, plate beamsplitters can include a thin, flat glass plate that has been coated on the first surface of the substrate, and typically feature an anti-reflection coating on the second surface to remove unwanted Fresnel reflections. Plate beamsplitters are often designed for a 45° angle of incidence.

8 8 FIGS.A andB 828 820 822 824 800 As shown in, the beamsplitter panelseparates the volume of the first compartmentinto a first regionand a second region. This angle is selected to offer the system a critical benefit of eliminating unnecessary lenses, thereby reducing complexity and cost, while at the same time improving ease of maintenance and overall reliability and performance. In other words, the 45-degree excitation angle was selected to decrease the number of parts, as this orientation would place the laser-PCB (printed circuit board) directly on top of the main-PCB, and still provided a desirable decrease in excitation relative to scattered radiation. Thus, in some embodiments, a Raleigh filter with an optical-chamber-integrated 180-degree co-linear setup is used for the second spectrometer. It can be appreciated that the use of fewer lenses streamlines the system, reducing potential optical losses and misalignments.

828 820 820 860 852 826 860 820 802 862 860 In some embodiments, as a result of the experientially-determined curated angle selection whereby the beamsplitter panelextends diagonally from a first corner of the first compartmentto an opposite and diagonal second corner (“kitty corner”) of the first compartment, each of the two regions can have a substantially triangular cross-sectional shape (i.e., isosceles triangle 90-45-45 degrees) across the horizontal plane, and/or an overall triangular prism three-dimensional shape. Along the first sidewallthat is adjacent to the first endand extends across the length of both compartments, a laser diode modulecan be mounted on the portion of the first sidewallthat extends along one side of the first compartment. The housingfurther includes a second sidewallruns parallel to the first sidewalland/or provides another wall that encloses the compartments provided within the spectrometer.

In different embodiments, a laser printed circuit board incorporated into the apparatus is capable of turning the laser on, off, and adjusting the power of the laser, allowing advanced users to fine-tune operational settings for their own purposes. In addition, in contrast to conventional Raman spectrometers that have incorporated diodes of lower excitation wavelength (e.g., a 785 nm 15 mW diode), in different embodiments, a more powerful and efficient 808 nm, 300 mW or 350 mW VCSEL diode can be incorporated in the present instrument. Modifying the excitation wavelength to a more powerful and efficient 808 nm 300 mW or 350 mW Vertical-Cavity Surface-Emitting Laser (VCSEL) diode enhances cost efficiency and signal quality. This diode has significant increases in performance, and decrease in cost, when compared to traditional diodes. In one embodiment, an 808nm wavelength allows the benefits of a higher wavelength while still remaining low enough for use as a traditional silicon detector. In some embodiments, the optical hardware can be calibrated with a Mercury-Argon laser.

In one preferred embodiment, the device can alternatively or also include an 804 nm 250 mW CWL laser. This type laser can provide a central point of emittance rather than being configured in an array that can in some cases result in an erratically focused laser. Similarly, in some embodiments where a VCSEL is used, it can also be configured as a focused emitter rather than an array.

802 820 830 830 840 820 840 830 Returning to the larger housing, it can be observed that the other end of the first compartmentis disposed directly adjacent to a slit portion. On the other (opposing) side of the slit portionis a second compartment. Thus, the first compartmentand second compartmentare spaced apart but also joined, connected, or bridged together via the slit portion.

840 842 860 840 826 920 840 920 9 FIG. In different embodiments, the second compartmentincludes an interior spacein which a transmission grating can be installed, and where the output of the scan is directed, toward the portion of first sidewallthat is associated with/encloses the second compartment(adjacent the laser diode module). Moving now to the view of, gratingis shown installed in the second compartment. In some embodiments, this gratingcan include a reflective diffraction grating, rather than a transmission grating to provide better efficiency and durability than more conventional reflection gratings. In some embodiments, the grating is a transmission grating that is adjustable, for example via a micro step-motor and/or worm gear.

9 FIG. 940 828 1 2 1 2 1 3 1 3 982 984 982 986 also provides a view of dichroic beamsplittermounted on/in the beamsplitter panel. In different embodiments, the beamsplitter is oriented diagonally across the compartment so that an angle Ais substantially equal to angle A. In some embodiments, angle Ais 45 degrees and angle Ais 45 degrees. In addition, angle Aand angle Acan be substantially equal. In some embodiments, angle Ais 45 degrees and angle Ais 45 degrees. Thus, it can be understood that in different embodiments, the dichroic beamsplitter is oriented along a second axisthat is diagonal, at a 45-degree angle, relative to a first axisalong which the collection lens is oriented (e.g., the lateral direction). The second axisassociated with the dichroic beamsplitter is also diagonal, at a 45-degree angle, relative to a third axisalong with the laser diode is aligned (e.g., the longitudinal direction).

In some embodiments, either or both of the adjustable transmission grating and dichroic beamsplitter are modular. Such a feature allows the user to readily remove and replace, and/or adjust either or both of the transmission grating and dichroic beamsplitter in cases where a lower or higher wavelength laser is desirable. For example, pockets or receiving slots or openings can be provided in the housing that can be left open for part retainers. These retaining structures can be configured to slide into the pocket securely, while the interior of the retainer would be customized for the particular component that is being added or swapped. These pockets could, in different embodiments, either be accessible after removing a screw from a faceplate, or a whole-body sleeve, that keeps the retainers in place, to ensure ready access to portions of the device where component access is more desirable.

940 862 826 960 960 Behind the dichroic beamsplitter, on the second sidewallopposite to laser diode module, a laser dump (not shown) can be provided at a dump site. In some embodiments, the dump sitecan also couple as a port for a calibration laser.

9 FIG. 930 860 930 950 860 920 950 920 930 also illustrates an example of a camera and PCB module (“camera”)that is connected or mounted to face toward the first sidewall. In some embodiments, the camerais positioned and oriented to captured the output of a filterinstalled on first sidewall, substantially facing the grating. In different embodiments, filtercomprises a small Raleigh filter that is incorporated between the gratingand photodetector (camera) to ensure precise photon placement and enhance signal quality by filtering out visible and excitation wavelengths. This allows the instrument to operate with a smaller filter, thus decreasing costs. The design parameters described herein—including angle adjustment, lens optimization, wavelength adjustment, grating enhancement, and filter integration—have shown markedly higher device efficiency, lowered costs, and enhanced data quality, resulting in a more robust and cost-effective system. The overall arrangement and reduction in components in the disclosed lens design also enable the correct alignment of the parts to be maintained even when the instrument is being “disturbed” significantly (bumped, jostled, dropped, hit, etc.) in contrast to conventional spectrometer instruments, which are highly sensitive to jarring motions.

950 930 930 Once the signal passes through the Raleigh filter, the cameracan communicate its captured data (e.g., imagery) to the mobile device, for example through a USB-C connection, lighting port, or other data transmission port. in some embodiments, the cameraworks in conjunction with a raspberry pi server, while communicating through the USB-C or lightning port or other charging, data transfer, and communication port provided with the selected mobile device.

930 As will be discussed further below, it should be understood that at the time a sample is taken using the proposed instrument, the Raman scattering may not be displayed or otherwise presented for the human eye to see. Instead, in different embodiments, the information that is collected is captured by the onboard instrument camera, which is configured to translate the signal peaks and troughs through RAW data image processing. This is possible because the grating angle is calibrated with a known spectrum. Using these optical pieces in the configuration illustrated herein provides the average angle of the grating with the highest efficiency possible. While other angles are possible, the current embodiment utilizes the highest efficiency angles based on the blaze wavelength of the grating. The calibration laser (which emits very specific wavelengths across the spectrum) can be used to calibrate where these wavelengths are landing, which will allow the instrument to accurately report where the peaks in measured samples are actually landing.

9 FIG. 10 FIG. 800 800 From the perspective of, the route that light can travel through the second spectrometeris more readily discernable. For purposes of clarity to the reader, an example of this pathway is schematically illustrated inby reference to a top-down view of second spectrometer. It can be appreciated that this pathway represents a flow of the general operational path for the proposed apparatus.

In different embodiments, the user can initiate testing by use a bottle/vial/substrate/slide to collect a sample (e.g., fuel or other bio/chemical compound), and fill to the desired sample level. In some alternate embodiments (not shown here), the bottle cap can be a consumable incorporating a filter or a SERS substrate. In such cases, the user can hold the bottle upside-down for a few seconds to wet the bottle cap, and then remove the bottle cap from the bottle and insert it into the Raman instrument's sample holder for analysis. The cap can further include a barcode for easy identification.

826 As noted earlier, laser diode module, for example including a laser diode or DPSS (diode pumped solid state) laser can be used to provide the light source. A diode-pumped solid-state (DPSS) laser is a type of solid-state laser that uses a laser diode to pump a solid gain medium, while a laser diode is a type of laser that uses electrical bias to inject charge carriers into a depletion region. In different embodiments, two or more laser diodes can be “stacked” to allow the user to select a different beam type. Dichroic filters could be selected, and corresponding laser wavelengths, and screw them (stacking multiple, if desired), onto the side of the housing with the laser dump. The light beam when activated or switched on can travel into the housing of the instrument where it can pass through a laser focusing lens.

1026 940 810 940 826 960 826 10 FIG. As depicted by the red line emerging from a laser focusing lensin, extending in an “Easternly” direction, the focused laser beam can then arrive at the dichroic beamsplitteror filter (e.g., interference filter) that selectively permits light of a smaller range of colors to pass while reflecting other colors. As the dichroic filter is oriented at 45 degrees relative to both the laser's axis of output and the collection lens, the filtered light is routed or “bounced” at a substantially right angle toward the collection lens(e.g., as represented by the continuation of the red line moving from the filter in a “Northern” direction). In addition, the relative positioning of dichroic beamsplitterand laser diode moduleallows for the incorporation of effectively situated laser dump (e.g., installed at dump site). Thus, spurious emissions that were removed/passed by the filter can continue to travel forward (“East”) and be harmlessly dissipated via a laser dump provided in the housing directly opposite to the laser diode module. In some embodiments, a mirror and/or a fiber optic input can be installed where the laser dump is located.

810 1 In different embodiments, the collection lenscomprises a physical component with an “electronic eyeball” that is able to see more colors than the human eye can, and is able to distinguish individual colors very precisely. The lens can direct and strike a sample with a beam of light, and receive a visual answer from the sample that is at a distance Dfrom the collection lens.

810 1 810 810 4 7 FIGS.andB In different embodiments, the collection lensis positioned so that a sample carried by a vial inserted into the adjacent holder (e.g., see) is consistently at a 20 mm distance (as D) from the collection lens. In different embodiments, the distance could be correlated to the focal length of the collection lens which could be 15 mm diameter, 20 mm focal length, and comprising an uncoated double concave lens. In other embodiments, the center of the vial (in which a sample can be held) when inserted into the receptacle can be between 15 mm and 25 mm from the collection lens. It should be appreciated that the importance of the focal length can directly correlate to the signal acquisition: for example, the smaller (more focused) the laser is on the sample, the more Raman signal from the sample will make it back to the detector. Furthermore, with respect to the vial, it can be placed in the vial-holder so that the focal point will be relatively in the middle or center of the vial, thereby providing a degree of variability in the focal point that can help accommodate the effect of aberrations in the material of the vial itself on the readings. The Raman signal can then follow the inverse pathway back to the optical components. Thus, the use of a vial holder can offer the user more security and repeatability than a “point-and-shoot” method, which allows for greater flexibility, but which also requires the user to accurately place the spectrometer the correct distance away from the sample. In some embodiments, the device can be part of a kit of parts that includes multiple types of vial-holders that can be of varying lengths so the user can swap out one holder for another to adjust the relative position of the vial from the optical elements, as desired.

810 940 830 840 920 920 Once the visual response is detected at the collection lens, it can be reflected backwards (to the “South”), unaffected as it passes through the dichroic beamsplitter, and travel through a slit (e.g., one or more apertures through which the beam passes) in slit portion, until arriving into the second compartmentin which gratingis provided. In some embodiments, the gratingcan be slotted or inserted into the chamber and/or removed when desired (i.e., modular gratings).

840 840 920 950 930 11 FIG. In different embodiments, the second compartmentcan be angled to take advantage of the blaze angle that is unique to each grating. Though the drawings depict the second compartmentin this case as being substantially rectangular, other embodiments (e.g., see) can include a rounded outer sidewall. Upon striking the grating, the light can be divided into a spectrum of spatially separated wavelength components that are directed through Raleigh filterand to the apparatus' photodetector/camera(shown installed on the “Western” side of the housing). This visual data is captured as a JPEG or other image data that is then transferred over a wired or wireless connection to the mobile computing device, where it is received at the associated processing application for analysis. Alternatively, raw data from the photodetector can be transmitted to the phone and translated into any image format desired.

It should be understood that the use of directional terms “North” or (N), “East” or (E), “West” or (W), and “South” or(S) are simply for purposes of ease-of-reference for the drawing as presented, and are not intended to restrict the embodiments in any way. In other words, in other embodiments, the components can be arranged at locations relative to one another that permit variations from these terms.

11 12 FIGS.and 11 FIG. 8 FIG.A 1100 800 1108 810 800 1160 1100 1170 1120 1130 1150 For purposes of illustration, an alternate embodiment is shown in. In, a third spectrometer apparatus (“third spectrometer”)with the same scanning components described with reference to second spectrometerinis depicted, but installed in a housingthat differs in geometry from the housingof the second spectrometer. More specifically, it can be observed that while a first compartmentof the third spectrometerremains substantially rectangular prism-shaped, the shape and size of a second compartmenthas been modified to permit variation in the orientation of a gratingrelative to both slit portionand Raleigh filter.

9 FIG. 1124 1100 1140 1140 1110 1110 1140 1124 1162 1124 As described earlier with respect to, when a scan is initiated, a laser diode moduleof the third spectrometercan emit a focused laser beam, which strikes a dichroic beamsplitterthat selectively permits light of a smaller range of colors to pass while reflecting other colors. The dichroic beamsplitteris also oriented at 45 degrees relative to both the laser's axis of output and collection lens, so that the filtered light will be routed or “bounced” at a substantially right angle toward the collection lens. In addition, the relative positioning of dichroic beamsplitterand laser diode moduleallows for the incorporation of effectively situated laser dump (e.g., installed at a dump site). Thus, spurious emissions that were removed/passed by the filter can continue to travel forward (“East”) and be harmlessly dissipated via a laser dump provided in the housing directly opposite to the laser diode module. In some embodiments, a mirror and/or a fiber optic input can alternatively or additionally be installed where the laser dump is located.

12 FIG. 1110 1230 1220 1240 1210 2 1110 1210 1100 Referring briefly to, in different embodiments, the collection lenscan be positioned so that a samplecarried by a vialsnugly inserted into a receiving receptacleof a modular sample holderis consistently at a distance Dfrom the collection lens. The modular sample holdercan be connected and disconnected from the exterior surface of the third spectrometer, for example via rails, sliding guides, slots and hooks, screws, magnets, etc. that when connected ensures the relative position of the two devices is consistently achieved when one sample holder is swapped out for a different sample holder (e.g., with a different receptacle or slot for holding other substrates/samples). In different embodiments, the distance between the sample and the collections lens when the sample holder is connected could be correlated to the focal length of the collection lens which could be 15 mm diameter, 20 mm focal length, and comprising an uncoated double concave lens.

11 FIG. 1110 1140 1130 1170 1120 Returning to, once a visual response is detected at the collection lens, it can be reflected backwards (to the “South”), unaffected as it passes through the dichroic beamsplitter, and travel through slit(s) (e.g., one or more apertures through which the beam passes) in slit portion, until arriving into the second compartmentin which gratingis provided.

1170 1182 1180 1180 1102 1160 1130 1104 1170 1182 As noted earlier, in different embodiments, the second compartmentcan be angled to take advantage of the blaze angle that is unique to each grating, including a rounded exterior wall that increases the interior space available within the compartment. More specifically, a first sidewallis now bent distally outward from the center of the device rather than remaining substantially linear/planar. Furthermore, a second sidewallballoons outward. Thus, the second sidewallcan include a planar portionthat extends between the first compartmentand the slit portion, before continuing in a curved direction to provide a bulging portionthat wraps around the second compartmentuntil arriving at the first sidewall.

10 FIG. 11 FIG. 920 1 830 1120 2 1150 1 2 1150 1132 In this way, the grating's orientation and position relative to the other components can be significantly adjusted to improve performance. For example, while inthe gratingwas oriented at a relatively narrow first angle Arelative to the slit portion, in, the gratingis oriented at a relatively larger second angle Athat receives a broader array of light as it arrives and better distributes and reflects the light as it is reflected to the Raleigh filter. In some embodiments, the first angle Ais less than 45 degrees, and the second angle Ais greater than 45 degrees. In different embodiments, the visual data received at the Raleigh filtercan then be captured by a camera moduleas a JPEG or other image data that is then transferred over a wired or wireless connection to the mobile computing device, where it is received at the associated processing application for analysis. Alternatively, raw data from the photodetector can be transmitted to the phone and translated into any image format desired.

(A) Laser: 808 nm wavelength, 804 nm wavelength, 785 nm wavelength, 180-degree co-linear excitation angle. As conventional designs required precise alignment in a controlled lab environment and were prone to disturbances (thereby making them unreliable for field use), the disclosed embodiments incorporate a co-linear setup that is far more robust and resistant to disturbances, allowing multiple points along the laser path within the sample to be used for signal collection. This flexibility substantially enhances reliability and performance across a wide range of rugged environments. The co-linear setup allows the device to take advantage of the surrounding structures of the spectrometer for mounting purposes. Any disturbance that occurs to the surrounding structure will result in the same disturbance to the laser and optical pathway, allowing them to remain parallel. In other setups where the laser is not on the same optical pathway, any vibrations through the device would result in a much more exaggerated divergence of the laser from the focal point of the collection lens. This angle uses the same pathway of excitation and collection, which results in an exponential increase in signal collection. A 0 degree or 180-degree would use this pathway. Any other angle would result in only a single point of the laser that would encounter the focal point of the collection lens. (B) Modular Grating: Robust ruled reflection grating with 1200 lines/mm and a 900 nm blaze wavelength can be provided. 1200 lines/mm is one of the higher typical specifications in for gratins, and the higher the number the greater optical resolution of the diffracted wavelengths. The 900 nm blaze angle corresponds to the average wavelengths that we are interested in measuring, and this number typically results in a higher optical efficiency. This results in a much more efficient setup than a grating with a higher, or lower, blaze angle. Conventional gratings are lightweight but fragile, unsuitable for demanding environments. The proposed apparatus' inclusion of ruled reflection grating is durable and optimized for high-stress (e.g., military) use, capable of withstanding structural stress while maintaining high diffraction efficiency in the NIR range. In some embodiments, the grating is readily replaceable for users desiring to modulate their diffraction targets, whereby a first grating can be slid out of its slot in the housing and a different, second grating slid back into the same slot. (C) Lenses: The collection lens is expanded to allow for a 15 mm beam, enhancing signal strength significantly. The lens could be 15 mm diameter, 20 mm back and front focal length, double concave lens. Every point of the laser, from the collection lens to the focal point (20 mm), will result in a point of radiation that can be collected. The greater the lens diameter, the more radiation that can be collected, which will result in a greater Raman signal. Most Raman analyzers will utilize a collection lens, and subsequent beam size, of <6 mm. This size is larger than those provided in conventional instruments and ensures improved signal collection. The proposed lenses are made of robust glass, providing superior optical clarity, durability, and resistance to environmental stress. In contrast to conventional lenses, that limit performance and durability, the larger lens support rugged and precise applications. (D) Filtering: Bandpass filtering can be provided via two filters—one for reflecting the excitation laser (˜90% reflection efficiency) and an RG850 filter directly before the optical detector, with preliminary results showing detectability ˜150 cm^−1. The proposed setup significantly reduces interference from the 808nm excitation laser, improving signal isolation and overall performance. 1 in a million interactions of the excitation laser could result in a photon corresponding to Raman radiation. As a result, the amount of light hitting the photodetector without filtering could result in the majority being the excitation laser. This will typically wash out the Raman signal, resulting in an output where users will only be able to see the excitation frequency. Bandpass filtering is then employed, thereby reflecting this specific frequency, which will block the laser and allow focus on the Raman signal. (E) Modular Detector (Camera): An off-the-shelf, modular, low-cost photodetector can be employed. Rather than rely on a camera incorporated in the smartphone for detection—which can be inherently expensive and required precise calibration to the phone in use—the proposed apparatus incorporates its own photodetector that can serve as a phone peripheral, allowing compatibility with any phone without requiring precise calibration. This approach reduces cost and complexity while increasing flexibility and usability. Multiple available models of photodetectors could be utilized. Non-limiting example photodetectors are the Sony imx219, imx708,imx462, or other image sensors compatible with Raspberry Pi cameras. Specifications of a photodetector could include a quantum efficiency (QE) for the wavelengths the users are interested in measuring (i.e., sensitive enough to detect those frequencies). For example, QE is expressed as a percentage, so that a QE of 75% means that 75% of photons are converted into electrons. Depending on the specific application for which the spectrometer apparatus is being used, a modular photoreceptor can be selected that has different semiconductor materials, thereby affecting the spectral response (which varies depending on wavelength). In different embodiments, a first photoreceptor can then be detached and a different, second photoreceptor attached as desired by the user. (F) Modular Sample Holder: In different embodiments, a sample holder can be provided as an integrated part of the apparatus. In other cases, the sample holder can be configured with an attachment or connection/disconnection mechanism by which it can be removably mounted to the sidewall of the apparatus in which the collection lens is installed, permitting swapping in and out of modular sample holders configured for different vial sizes and shapes and/or microfluidic slides of varying dimensions. In addition, in some embodiments, the sample holder can feature an open well to prevent residual fuel from pooling. In this way, the holder facilitates is compatible with the adjoining compact spectrometer, allowing for easy insertion and analysis of samples. In different embodiments, the sample holder can further include provisions for integrating a SERS substrate to enhance trace contaminant detection and quantification. Furthermore, real-world testing with the sample holder involving various disposable pipettes showed there was no interference with the Raman spectra when transferring fuels. As a non-limiting example, a concave reflective diffraction grating, which would focus all of the diffracted wavelengths directly towards the photodetector could be used. Incorporating a rotating Powell lens can allow the system to cast the excitation laser as a line, as well as collect the radiation directly from that line. This would bring the performance of the device near to that offered by a conventional hyperspectral Raman spectrometer, by allowing the device to pick out specific particles and list their corresponding spectra. (G) Housing: In different embodiments, the main housing could be composed of a rigid, robust, resin polymer mixture that is rated to withstand 450 degrees Fahrenheit. The individual components can be separated from the housing using an insulator of the same material, allowing for custom insulators to be made for each type of device the user would like to mount. The devices, and insulators, lend themselves very well to modularity and customization, and can be swapped out by hand. In some embodiments, the devices, and insulators, are mounted to the main housing by small metric screws, facilitating user access and repairs/replacements when desired. For purposes of reference, in different embodiments, each of the proposed spectrometer apparatuses described herein can comprise scanning components/devices associated with the following material specifications:

13 FIG. 110 120 232 112 As noted above, in different embodiments, the apparatus is configured for use as an accessory with a smartphone or other mobile computing device, making it highly appropriate and practical for repeated use in the field. In, a real-world test scenario is illustrated in which the first spectrometeris connected to mobile device, and a sample contained in the first vialhas been inserted into and received by receptacle. In different embodiments, the scanner can detect contaminants and live microbial pathogens including bacteria, mold, and fungi, using Raman Spectroscopy.

In some embodiments, the spectrometer's photodetector (camera) can be associated with a processor/memory component that can include or be otherwise connected to a mobile computing device adaptor that can connect directly into the power/data port of the device. For example, the images are transmitted to a smartphone, and received by a mobile phone application employing a trained Deep Neural Network (see below). The instrument, when connected to the smartphone, thereby provides users with a real-time adaptive, easily trainable, artificially intelligent handheld instrument for chemical and microbial point detection in the field.

In different embodiments, the apparatus can be plugged into a mobile phone, and draw power from the phone to an onboard battery, which is used to power the CPU, photodetector, and laser diode. In different embodiments, the processor could handle the controlling signal to and from the phone, and translating that into something that the photodetector can understand, and vice versa. As a non-limiting example, iOS® and Android® phones and tablets generally have one primary data port that is used for charging and for connecting to accessories. Thus, in different embodiments, the spectrometer apparatus can be configured with a modular adaptor that can provide a secure, stable connection to Type USB-C, micro-USB, and Lighting® connections, or other data port type that is developed.

122 In some embodiments, the apparatus can include multiple types of male adaptors (e.g., as part of a kit) for connecting the apparatus to a desired mobile computing device data port. In one embodiment, an adaptor can be rigid and resilient enough to support the mounting of the apparatus on the smartphone. In different embodiments, the adaptor can include a mechanism for decoupling the apparatus from the smartphone, such as a squeeze button or lock/unlock mechanism. Once the apparatus is linked to the data port, the spectrometer can be configured to operate as a “plug and play” device so that data collection can be performed with no further mechanical adjustments between the two parts. Additionally, a package on the device could allow users to plug it into the device and install the app from there. For example, the apparatus can be pre-calibrated to work with any type of smartphone, a feature that is possible due to the incorporation of an onboard photodetector (rather than relying on the smartphone's camera). In some embodiments, the apparatus can be powered through the data port. In one example, the apparatus can receive instructions or other control signals from the smartphone through the data port that allow the user to turn on the detector, turn off the detector, initiate a collection event from a sample, etc. These options can be provided via a GUI for a mobile application installed locally on the smartphone (e.g., a “Contaminant Recognition feature” provided via the app). In other embodiments, the apparatus itself can include a mechanical power on/off switch.

1300 122 110 122 1322 1322 1322 1322 1310 1320 1322 An example of a first graphical user interface (GUI)associated with the appis shown in which incoming data collected via the first spectrometerhas been received by the appand processed to generate and display an interactive spectral graph. In this example, the spectral graphincludes a plurality of “peaks” that correspond to the highest points on the graph, indicating the wavelengths or frequencies where the intensity of a signal is the strongest, signifying the dominant components within the signal data being analyzed, as visualized on a spectrum graph. In some embodiments, the spectral graphcan include features whereby a user can increase or decrease magnification to view with more clarity the peaks, and/or interactive regions such that a user can tap to view the chemicals that have been identified and their relative amounts in the sample. In different embodiments, the first GUIcan further include selectable options such as a first optionto capture and remove the background, and a second optionto “subtract”, or isolate specific features or patterns in the data. Other options can also be provided to enable navigation from the first GUIto other menu items.

13 FIG. It should be appreciated in observing the parts ofthat the proposed systems are designed to further both structural robustness and device miniaturization. The proposed design is directed toward repeated use in high-stress, rugged environments (e.g., military applications) and rely on an integrated optical detector. The number of parts in the disclosed embodiments are minimized by leveraging dual-use components, maintaining affordability, and enabling a compact and miniaturized design. In contrast to conventional devices designed for laboratory use that are unable to withstand impacts or other stressors, the proposed embodiments offer a rugged, reliable, and portable system suitable for field operations, all while keeping costs significantly lower than those associated with conventional Raman spectrometers.

In different embodiments, the GUIs described herein can rely on open-source software and systems. In contrast to proprietary operating systems, which restricts their use to original equipment manufacturers' (OEM) hardware, and limits the use of the detector in the field, a smartphone-based Raman spectrometer as described herein can incorporate a modular and open-source approach. In one example, the embodiments can be configured to work in conjunction with an application providing users with a relatively simple GUI (e.g., under the SMC-S-023, Human Computer Interface Design Criteria for GUI development). In different embodiments, the software app, which can be built on an Android® Operating System (OS) or Apple® iOS, can employ an iterative testing approach with real end-user feedback to provide an intuitive use process requiring little to no prior instruction or manuals.

14 18 FIGS.- 14 15 16 FIGS.,, and 14 FIG. 15 FIG. 1400 1410 1420 1430 1500 1510 1520 1530 1540 1510 1600 Some examples of GUIs for the app are depicted in.show an example sequence of GUIs by which a user can navigate from an initial test start screen to an active testing stage. More specifically, a second GUIofis directed to performing a basic fluid source testof a sample submitted via cuvette. In some embodiments, the GUI can further include a first selectable optionto initiate a new scan of a sample, as well as a second selectable optionto view recent/previous test data. Moving to, in some embodiments, once a user has initiated a scan, a third GUIcan be presented by which one or more chemical profile types are displayed for selection. In this specific case, the app is directed toward measuring contaminants in fuel; thus, the profile types include a first profile(e.g., fuel for a jet/aircraft), a second profile(e.g., fuel for a first type of tank), a third profile(e.g., fuel for a second type of tank), and a fourth profile(e.g., fuel for an all-terrain vehicle). In other embodiments, the app can be configured to present a wide range of other profiles, related to fuel and non-fuel compounds. For purposes of this scenario, the user selects the first profile. In response, the app can cause a scanning operation at the connected spectrometer-dongle to begin, as illustrated by a fourth GUI(“Test in Progress”).

1700 1710 1720 1730 1740 1750 1760 Upon completion of the scan, the app can automatically process and analyze the data captured by the spectrometer, and display a results screen. In some embodiments, this screen can immediately alert the user as to the presence of one or more pre-designated contaminants, or whether the sample has been determined to be uncontaminated. For example, in a fifth GUI, a first test dashboard is presented. This dashboard can include an identifier label/file namefor the sample for reference (automatically generated or inputted by the user), and data associated with the test, including but not limited to: (a) a summary message(“PASS: Contamination not detected”); (b) date, time, and location of the test; (c) specific chemicals or contaminantsthat were searched for in the sample and whether they were present/detected (in this case, PASS indicates no such contaminant was detected); (d) which profile was selected; and (e) a spectrographof the data.

18 FIG. 1800 1820 1810 1830 1840 1860 1800 1850 In contrast, by reference to, an alternate test scenario result is shown where contaminants were in fact found in the sample, triggering presentation of a sixth GUI. In this case, a second test dashboard is presented. This dashboard can include an identifier label/file namefor the sample for reference (automatically generated or inputted by the user), and data associated with the test, including but not limited to: (a) a summary message(“FAIL”) including a targeted recommendation (“Based on regulation 321 it is recommended to confirm contamination with SERS testing”); (b) date, time, and location of the test; (c) specific chemicals or contaminantsthat were searched for in the sample and whether they were present/detected (in this case, FAIL indicates a particular contaminant was detected); and (d) a spectrographof the data. Furthermore, in some embodiments, in response to the FAIL decision outputted by the system, the sixth GUIcan include a quick-optionto move directly to initiating the SERS test, as recommended.

19 FIG. In different embodiments, the GUI can include different profile systems within the device (e.g., end-user, chemist, and administrator, etc.). The end-user profile could have the lowest number of permissions, preset laser and limited functionality controls, while the chemist's view could allow the adjustment of the laser strength and exposure time. The chemist's view could also be restricted from changing the any software presets or uploading data to a Federated Learning Platform (e.g., seebelow). Furthermore, the administrator view could include unlimited access to the device's capabilities. The results screen would also depict an image of the spectra captured by the apparatus, as discussed above.

As presented herein, the proposed app interface is designed a simple, yet highly effective GUI tailored for minimally trained operators. This interface can thereby facilitate real-time fuel and propellant analysis by providing clear, step-by-step prompts and intuitive displays of critical parameters. The GUIs can present results with clear pass/fail indicators, ensuring rapid decision-making without requiring deep technical expertise.

As introduced earlier, in different embodiments, the proposed system includes not only a robust, powerful apparatus, but enterprise software integrated into the mobile application that employs an intelligent Deep Neural Network (DNN) algorithm. In different embodiments, the algorithm is configured to apply signals processing techniques (e.g. wavelet transform or Continuous Wavelet Transformation (CWT)) to detect and quantify aviation grade Polyalphaolefin (PAO), sulfur compounds, hydraulic fluid, and microbial contamination in jet fuel samples, as well as a wide range of other aberrations in samples. Advanced signal processing techniques such as CWT provide a more effective means of analyzing transient fuel properties in real-time without extensive sample preparation. Unlike traditional global transform techniques, CWT allows localized, multi-scale analysis of fuel samples, improving precision and robustness in field environments. The use of AI-driven chemometric modeling further enhances analytical accuracy, reducing the need for human interpretation and minimizing operator training requirements. These advancements collectively enable the development of a deployable, low-cost spectrometer capable of real-time, high-accuracy fuel and propellant analysis, addressing the deficiencies of previously tested method. In some embodiments, the algorithm is configured to provide a jet fuel thermal stability indication.

For purposes of reference, some details regarding an embodiment of the proposed AI model are provided. In different embodiments, the system can include a deep neural network (DNN), feed-forward neural network (FNN), and/or convolutional neural network (CNN) that can be trained to classify different levels of fuel contamination using Raman spectra. In one testing scenario, the training datasets included samples with contamination levels of 1%, 0.5%, 0.25%, 0.125%, 0.0625%, and 0%, using Paragon® and Marathon® jet A fuel samples. The training process involved using machine learning algorithms to analyze the Raman spectra and develop a model capable of accurately identifying and quantifying contaminants. Additional data from external entities were incorporated into the training process to enhance the model's accuracy and reliability. The DNN was thereby trained to achieve high accuracy in detecting and quantifying various levels of contaminants in jet fuel by collection and prepare training datapoints with varying levels of contamination using datapoints obtained from real-world testing samples. The model performance was then validated using additional samples from various fuel providers. The DNN demonstrated a high degree of accuracy in classifying fuel contamination levels using Raman spectra, successfully detecting contamination levels at 1%, 0.5%, 0.25%, 0.125%, and 0.0625%. The DNN can be continuously updated and improved using new data and feedback.

25 25 FIGS.A andB In some embodiments, the resultant AI model is able to take the photodetector's captured colors and their corresponding brightness and intuit what the sample is made up of. In contrast to reliance on a database of entries, which would be static and require a significant consumption of memory and processing, the AI model can be trained to inherently understand that “ingredient” list and immediately recognize the contaminant(s). In different embodiments, the AI model can also or alternatively be trained to recognize unique speckle patterns to recognize a sample, as discussed below with reference to.

In different embodiments, the system can be further refined to expand upon the AI neural networks described above to enhance its predictive analysis capabilities for Raman spectral data. For example, by leveraging advanced algorithms, the system can be trained to trends and anomalies, predict molecular compositions, and simulate macroscopic fuel properties. These models can evolve and refine their predictive accuracy using a combination of experimental and synthetic datasets, enabling robust analyses in real-time field conditions. Key parameters, including electrical conductivity and freezing points, traditionally requiring direct measurement under controlled conditions to ascertain such properties, can then be estimated with precision by correlating Raman spectral features to molecular structures and properties. In some embodiments, this process involves refinement of existing neural network architectures to process Raman spectral data for identifying molecular patterns and predicting chemical properties. In different embodiments, this process can incorporate both experimental and synthetic datasets (see below) that are used to extensively train the model to ensure adaptability across various fuel types and environmental conditions. To strengthen model performance, an advanced anomaly detection algorithm can also be integrated into the network, enabling the identification of outlier patterns that could indicate contaminants or unexpected chemical variations. In addition, the system can include rigorous validation processes that simulate extreme environmental conditions, allowing the system to sustain accuracy across a wide range of operational scenarios, from extreme cold to arid climates. Advanced adaptive algorithms can also continuously calibrate the spectrometer in real time, mitigating the effects of environmental factors such as temperature and humidity shifts. Furthermore, integrated anomaly detection and automated corrective actions prevent system failures, while user training modules ensure operators at all technical levels can effectively utilize the system with minimal error.

19 FIG. 1400 1410 1420 1430 1450 Furthermore, referring next to, in different embodiments, the system can include, employ, or access a Federated Learning (FL) marketplacefor refining of the algorithm, employing a decentralized approach to training Machine Learning (ML) models. In such cases, raw data at the edge is used to train the model locally, which promotes and increases data privacy. In one example, the final model is generated via a shared learning pool, by aggregation of the local model outputs in different locations and/or facilities across the world (e.g., from a first local model, a second local model, and a third local model). A larger, global FL modelcan thereby be trained, for example, to detect new contaminants from the provided Raman spectra. The production of data across multiple use-cases (e.g., Commercial Airline, US Coast Guard, US Air Force, Navy, US Marine Corps, NASA, etc.) at different locations can be brought together to enable a collaboration by which the deep learning model continues to be trained and improved over time to detect contaminants in aviation fuel. Each organization can have their own data, and yet can be granted the benefits of a larger data resource that be used to train a model on all of the data without directly sharing data with each other, or with a central entity, ensuring privacy and data protection. The DNN model can thereby be trained in a federated manner, where each entity could train a model on its own data, and the models could be aggregated to produce a final model.

In different embodiments, each individual device could receive or download a copy of the most recent/up-to-date version of the global FL model network. For example, when one of the devices connects to the network, the system can determine whether there are any updates available, and if so, the updated weights and biases within/for the model network are exchanged (e.g., the user device sends theirs, and they obtain the updated one from the network). However, in such cases, confidentiality of user data would be maintained, so that no readable data is exchanged (e.g., spectra measured, sample data, etc.). In some embodiments, this process would be a periodic or as-needed update definable by or otherwise scheduled by the user, or occur in response to a manual request by the user.

By incorporating Federated Learning (FL) into field-level fuel and propellant analysis systems, their adaptability and analytical precision are enhanced while maintaining robust data security. By leveraging a decentralized ML platform, the system can continuously improve and refine its neural networks used for analyzing complex chemical profiles in real-world operational environments (such as, but not limited to, US military and commercial aviation sources), to generalize effectively across various fuel types and contamination levels. In different embodiments, the system employs secure FL protocols and lightweight model architectures that can operate efficiently in bandwidth-constrained and remote conditions, ensuring state-of-the-art performance and reliability. The FL can further incorporate secure protocols that anonymize, encrypt and transmit the results of the local models, allowing devices to transmit model gradients, weights and performance metric updates to a central server without compromising user privacy. In some embodiments, the system can incorporate over-the-air update mechanisms to enable automated, real-time improvements to the models. These updates shall ensure that the analytical system evolves continuously based on data from diverse sources, including jet fuel, diesel, synthetic fuels, and contaminants, while adhering to stringent privacy regulations.

As noted earlier, in different embodiments, the system can benefit from the use of synthetic datasets for refining of its neural networks. In one embodiment, simulated Raman spectra data can be generated to direct the performance of the neural network toward a particular target. Traditionally, spectra are analyzed directly against a database for a match. If there is no match, the sample is unknown. Every analyzer will show a slightly different spectrum for each sample, and a user would traditionally want to build their database using their own samples and analyzer. However, by utilizing synthetic mathematical approximations of spectra, the system can simulate how the analyzer would evaluate unknown samples. By utilizing a cascade of millions of mathematical approximations, the system can recognize what specifically is resulting in these spectra from specific molecules, and enable users to translate the data back and forth.

In some embodiments, an automated Density Functional Theory (DFT) algorithm can be created to generate artificial Raman spectra, enabling the simulation of real-world data for improved analysis. For example, the system can employ DFT to generate artificial spectra (which is a mathematical approximation) for individual molecules and employ Mass Spectrometry to determine molecular concentrations within specific samples. By constructing a mixture of artificial spectra using a linear concentration model, insights from experimental Raman spectroscopy can be obtained with AI-driven spectral translation. In one embodiment, the AI model is trained to correlate experimental spectra with their artificial counterparts, enabling efficient qualitative and quantitative spectral deconvolution. This method streamlines the interpretation of complex spectra and enhances the accuracy of molecular analysis.

In one example, the method of producing this synthetic spectral data can involve a process comprising: (1) generating millions of Mass Spectrometry (MS)-spectra; (2) converting the MS spectra to Raman spectra; and (3) Training of a Deep Learning Algorithm to recognize the millions of generated MS-spectra available, where these spectra include mixes of chemicals (bulk matrix) that the Raman spectrometer has not yet been able to identify. Such an approach would enable conversion of existing Mass Spectrometer and NIR spectra to a format that Raman could understand/process, allowing low cost, portable RAMAN spectrometers as disclosed to offer far broader and comprehensive use across industries where mass spectrometry dominates. In this way, the system allows for manufacturing of training data. For example, neat jet fuel can be obtained, and the target contaminants intentionally added to the fuel. These samples can then be excited with a Raman laser, and the spectra captured. A deep learning algorithm can classify the spectra and contamination. Once the base model is developed, future users will be able to use their own algorithms/models and provide the model updates through the federated marketplace. In some embodiments, the marketplace can include an aggregator that will push out new models to all users globally.

In different embodiments, the synthetic spectral data generation can revolve primarily or entirely around DFT (deep functional theory) calculations, as networks can learn how to simulate DFT calculations much more accurately than a purely algorithmic approach. As a general matter, the DFT-based approach involves the application of algorithms and analytical processes focusing, in modeling, a molecule at the quantum level, simulating how photons interact with it, and how the molecule changes based on these photonic interactions. There are numerous algorithms and analytical processes available for this. The system will include a custom-developed neural network to help make this process much more seamless and less calculation-intensive.

As described herein, Raman spectroscopy serves as the primary technique for capturing vibrational signatures of hydrocarbons and aromatic compounds in jet fuel samples, while dielectric property measurements add another layer of molecular characterization, particularly concerning electronic polarizability and phase transition behavior. The synergistic use of these techniques facilitates a novel approach for predicting macroscopic fuel properties through inverse computational spectra generation. By correlating the acquired Raman spectral data and dielectric properties with composition databases, the system can benefit from reference to reconstructed molecular profiles that serve as predictive indicators for critical fuel performance metrics. For example, to estimate the net heat of combustion, a computational model will be trained on historical Raman spectral data and calorimetric measurements, establishing empirical correlations between molecular structures and energy release characteristics. The dielectric constant, when analyzed in conjunction with Raman-derived hydrocarbon chain distributions and aromatic content, would enable prediction of the freezing point by capturing phase transition behaviors. Additionally, flash point and smoke point predictions can be derived from spectral signatures of long-chain hydrocarbons, aromatics, and molecular weight distributions, facilitating accurate assessments of volatility and sooting tendencies. The identification of oxidation-sensitive functional groups, including peroxides, carbonyls, and unsaturated hydrocarbons, through Raman spectroscopy provides direct insight into thermal oxidation stability.

20 31 FIGS.A- 20 20 FIGS.A andB 8 FIG.A 11 FIG. 21 FIG. 2000 800 1100 2000 2000 2182 2014 2000 For purposes of illustration, an alternate embodiment of a spectrometer apparatus that can include some or all of the features described above is now presented in. In, a fourth spectrometer apparatus (“fourth spectrometer”)is depicted that includes several of the same optical components described earlier with reference to second spectrometerinand third spectrometerin. The fourth spectrometeris also man-portable and compact, being sized and dimensioned with a form factor and weight (e.g., under 50 grams) designed to be held in a human hand or palm on its own, or carried as part of a smartphone/tablet via its click-and-secure connection to the smartphone or tablet device's data transfer port, enabling easy transport and sampling sessions at any location. In other words, the spectrometer is small enough that a person holding and/or carrying their phone in the palm of one hand could also carry the fourth spectrometerconnected to the phone—for example, in the form of a dongle (e.g., linked to the phone via a standard type connector/adaptordepicted in, protruding from the rearward side). However, as will be discussed below, the fourth spectrometercan also include one or more components that differ and can significantly enhance its spectral resolution, operable range, and number of spectral channels.

20 20 FIGS.A andB 20 FIG.A 2000 2000 2002 2002 2008 2018 2002 2094 2008 2024 present an introduction to the fourth spectrometer. In, the fourth spectrometeris shown fully assembled in a housingthat securely encloses an assembly of components within. In some embodiments, the housingcan include apertures, holes, or other openings to allow data from a sample to be captured via a collection lens, or a laser dump can be installed (e.g., a first openingand a second opening), etc. In some embodiments, the housingcan include a substantially rectangular prism-shape. In different embodiments, the shape can further include protruding portions such as a first protruding portionthrough which the first openingcan be formed with a greater thickness, increasing protection of the lens within and facilitating sample capture, and a second protruding portionthat can improve stability of the device.

2000 2086 2082 2084 2082 2000 2010 2014 2010 2014 2084 2016 2012 2000 2016 2012 For purposes of this application, the fourth spectrometerand each of its components thereof can be described by reference to a vertical axis, a longitudinal axis, and a lateral axis. The term “longitudinal,” as used throughout this detailed description and in the claims, refers to a direction extending along the length of a component (from the rear of the component to the front), in this case aligned with the longitudinal axis. For example, a longitudinal direction of the fourth spectrometerextends from a forward sideto a rearward side. The term “forward” or “front” is used to refer to the general direction which lies toward the forward side, and the term “rearward” or “back” is used to refer to the opposite direction, i.e., the direction which lies toward the rearward side. In addition, the term “lateral direction,” as used throughout this detailed description and in the claims, refers to a side-to-side direction extending along the width of a component (i.e., parallel to lateral axis). In this case, the lateral direction may extend between a first sideand a second sideof the fourth spectrometer, with the first sidebeing disposed on one side of a longitudinal midline, and the second sidebeing disposed on the opposite side of the same longitudinal midline.

2086 2000 2022 2024 Furthermore, the term “vertical,” as used throughout this detailed description and in the claims, refers to a direction generally perpendicular to both the lateral and longitudinal directions (i.e., aligned with vertical axis). For example, in cases where a component is disposed on a ground surface, the vertical direction may extend from the ground surface upward. It will be understood that each of these directional adjectives may be applied to individual components of the fourth spectrometer. For convenience, the term “upward” will refer to the vertical direction heading away from a ground surface (e.g., if the device were placed on the ground), while the term “downward” refers to the vertical direction heading toward the ground surface. Similarly, the terms “top,” “upper,” and other similar terms refer to the portion of an object substantially furthest from the ground in a vertical direction, and the terms “bottom,” “lower,” and other similar terms refer to the portion of an object substantially closest to the ground in a vertical direction. For example, a vertical direction may extend between a top sideand a bottom side.

2002 For clarity, the description also makes reference to distal and proximal directions (or portions) in the context of the spectrometer and its components. As used herein, the distal direction is a direction oriented away from the center and toward the outermost surface of the housing, while the proximal direction is an opposing direction that is oriented away from the outer housing and toward the center. In addition, the proximal direction can also be referred to as an “inward” direction, and distal direction can be referred to as “outward” direction.

2010 2014 2016 2012 2022 2024 2000 2000 It will be understood that the forward side, rearward side, first side, second side, top side, and bottom sideare only intended for purposes of description and are not intended to demarcate precise regions of the fourth spectrometerand its components thereof. Likewise, the first side and the second side and/or the top side and bottom side are each intended to represent generally two opposing sides of the spectrometer and each component, rather than precisely demarcating the fourth spectrometeror its components thereof into halves.

20 FIG.B 2000 2030 2000 2030 2074 2022 2010 2024 2014 2016 2012 2030 2072 In, for purposes of clarity, the fourth spectrometeris depicted without its outermost housing, thereby exposing its interior assembly. In different embodiments, spectrometer components can be mounted or installed within a chassisthat is disposed in an interior space of the housing of the fourth spectrometer. In one embodiment, some of the components can be arranged in a substantially linear or on-axis position relative to one another. In some embodiments, the chassisincludes a substantially rectangular prism shape, including a set of outermost sidewallsthat can substantially enclose or frame each of the top side, forward side, bottom side, and/or rearward side, as well as first sideand second side. In addition, in different embodiments, “behind” the sidewalls and formed in the interior of the chassisthere can be a plurality of slots (“slots”)or grooved sections or other joinery-type trench cuts, that can each sized and dimensioned to receive a corresponding optical module, described in greater detail below.

2030 2012 2030 2030 2096 2094 2002 20 FIG.B 20 FIG.A As an overview for the reader, the component modules installed in the chassisare described with reference to, where a top-down view of the device facing toward the (exposed) second sideis depicted. The full chassiscan in some embodiments include an overall shape that is substantially similar to the shape of the housing. In one embodiment, the chassisalso includes a corresponding chassis protruding portionthat can be snugly fitted into the slightly larger space formed by the protruding portionof housingthat was shown in.

2096 2094 2008 2004 2006 In addition, in different embodiments, chassis protruding portioncan include a chassis openingthat is aligned with the first openingso that a through-hole/transparent passageway between the outside environment and a collection lens (e.g., provided by a first lens moduleor a second lens module) can be maintained and light can travel without obstructions between a sample and the collection lens.

20 FIG.B 2004 2096 2010 2004 2022 2042 2030 2006 2004 2030 2014 2052 286 2058 In different embodiments, embodiments of the proposed spectrometer can include an optimized internal geometric configuration and assembly to streamline the device and enhance its performance. As shown in, first lens moduleis disposed adjacent to the chassis protruding portionformed along the forward side. In some embodiments, first lens modulecan correspond to the collection lens component. Nearest to the top side, adjacent to a top endof the chassis, is a second lens module. In addition, opposite to the first lens module, and adjacent to a portion of the sidewall of the chassisalong the rearward side, is a third lens module. Further “down” relative to vertical axisis a fourth lens module.

2004 2006 2052 2058 In different embodiments, each of the four lens modules can include individual lenses that are each secured in their own casing or block. In addition, in some embodiments, each of the four lens modules can be interchangeable with one another. For example, one or more of the lenses of each of first lens module, second lens module, third lens module, and fourth lens modulecan include a plano-convex lens (e.g., 8.0 mm diameter×10.0 mm focal length, NIR|Coated). A plano-convex lens is an optical lens with one flat surface (plano) and one outward-curving (convex) surface, giving it a positive focal length that converges parallel light rays to a focal point.

2030 2184 2030 2014 2052 2004 2054 2054 2060 2064 2054 21 FIG. In different embodiments, along an exterior of the chassis, a laser diode (see for example laser diodein) can be directed through one side of the chassis(e.g., the rearward side) so that, when activated, a laser can pass through the third lens module, which collimates the laser emission from the diode and then focuses the collimated beam across to the first lens module(e.g., via beamsplittersituated centrally or medially between the two components) and into the sample. For example, in different embodiments, the beamsplittercan be used to split the collimated laser beam between one or more sample testing locations and to reflect the collected Raman signal “downward” toward a pinhole module (“pinhole”)and onward to optical compartment. In some embodiments, the beamsplittercan comprise a cube beamsplitter (as shown in the present drawing) or a plate beamsplitter (as shown in earlier figures). In general, cube beamsplitters are constructed using two typically right-angle prisms, where the hypotenuse surface of one prism is coated, and the two prisms are cemented together so that they form a cubic shape. Furthermore, the beamsplitter can be dichroic, or in this case, non-polarizing and configured to split light into a specific R/T ratio while maintaining the incident light's original polarization state.

2004 2054 2058 2054 2056 2058 2058 2060 In some embodiments, the first lens modulethen collects or otherwise receives Raman-scattered light from the sample, which is passed back through the dichroic beamsplitterand re-directed orthogonally to the fourth lens module. The collimated Raman signal can pass or travel through the beamsplitter, which redirects the signal downward toward a filter modulebefore arriving at a fourth lens module. In one embodiment, the fourth lens modulecollimates the collected Raman signal, and then focuses the collimated Raman beam onward toward pinhole module.

2004 2004 2056 2056 2156 2158 2156 2156 2158 2158 21 FIG. 22 FIG. 21 FIG. In different embodiments, the second lens modulecan serve as a secondary or alternative collection lens site. In other embodiments, the second lens modulecan be removed and the site used as a laser dump and/or as a port for a calibration laser. In different embodiments, the Raman-scattered light first passes through a filter module(e.g., see) that can comprise one or more filters. For example, as depicted in, the filter modulecan include two or more separate filters. In, there are two filters, identified as a first filter componentand a second filter component. In some embodiments, the first filter componentcan be used to reject ambient visible light and partially attenuate the excitation wavelength by blocking wavelengths below, for example, ˜830 nm, while transmitting Raman signals above ˜850 nm. As a non-limiting example, in one embodiment, first filter componentcan include a SCHOTT RG850 Longpass Filter. For example, the filter can include a diameter of or around 12.5 mm and a thickness of or around 2 mm. In one example, the filter includes colored glass. In addition, in different embodiments, the second filter componentcan be used to suppress the strong excitation laser signal at 808 nm while allowing the surrounding Raman-shifted wavelengths to pass. As a non-limiting example, in one embodiment, second filter componentcan include an 808 nm Notch Filter. In one example, the filter can include a diameter of 12.5 mm, and an average optical density (OD) of around 4.0.

2056 2058 2060 2058 After exiting the filter modulethe signal can travel onward and pass through the fourth lens modulebefore arriving at pinhole module. In one embodiment, the pinhole is positioned at the focal point of the final plano convex lens of fourth lens moduleto spatially filter the Raman signal. The pinhole can, for example, block out-of-focus and/or stray light, allowing only the focused Raman signal to pass. The size or diameter of the pinhole may be adjusted to balance signal throughput and optical resolution. In one embodiment, the pinhole has a diameter of approximately 100 μM.

2060 2062 2062 2062 2062 2064 Once the Raman signal passes through the pinhole formed in the pinhole module, in some embodiments, the Raman signal can be received at and pass through an achromatic lens module. In different embodiments, the achromatic lens modulecan capture and collimate the transmitted Raman signal while minimizing spherical and chromatic aberrations. In one embodiment, the achromatic lens modulecan include a lens of 9 mm diameter×12 mm focal length, NIR II Coated. In some embodiments, the achromatic lens moduleensures a clean, well-collimated beam is outputted and then delivered to the adjacent optical compartmentfor spectral dispersion.

20 FIG.B 21 FIG. 20 FIG.B 2064 2024 2068 2068 2030 2000 2066 2064 As shown in, in different embodiments, the optical compartmentcan be disposed nearest to the bottom side, and include a distal endthat corresponds to the lowermost region of the assembly and will route the light signal to the photo detector or other sensor (e.g., see). In some embodiments, the distal endis flush with or protrudes slightly out of an opening formed in a bottom of the chassis. In addition, as will be described in greater detail below, the fourth spectrometercan include one or more metasurfaces (not shown in) installed in a dispersion unitformed in the optical compartment. These metasurfaces can serve as the device's dispersive element, and offer an alternative to the grating described earlier.

2000 With the incorporation of these metasurfaces, the proposed embodiments offer a significant technological shift forward from conventional spectrometers. For example, the embodiments described herein can replace costly optical components with engineered metasurfaces that bend and guide light at the nanoscale, eliminating bulk and expense without sacrificing performance. While traditional Raman spectrometers are massive, non-portable, and rely on a delicate alignment of lenses, gratings, and filters—each representing a potential point of failure and high-cost—the proposed devices introduce a metasurface architecture that integrates these functions into a single ultra-thin optical layer, reducing size and weight while boosting durability. Combined with custom built AI-driven noise reduction and spectral interpretation, the fourth spectrometercan be configured detect target signatures with sub-parts-per-million sensitivity, even in challenging real-world environments like a dusty farm, a humid clinic, or a bustling border crossing. These metasurface-based dispersion units will be discussed in greater detail below.

20 FIG.B 2076 2064 2078 2030 2030 2100 2000 2100 2030 As demarcated in, for ease of reference, the “upper” portion of the device that comprises all of the lenses and will be referred to as an optical routing portionand the remaining “lower” portion of the device (including the optical compartment) will be referred to as the optical dispersive portion. Collectively, the full chassisalong with the interchangeable modules installed in the chassiswill be referred to as a spectrometer assembly (“assembly”)of the fourth spectrometer. In different embodiments, a kit of parts for the spectrometer can include the assemblyinstalled in the outer housing, as well as a set of alternate modules that can be swapped for the modules currently slotted in the chassis, and/or one or more modular photo detectors that can be attached to the end of the chassis.

21 FIG. 20 FIG.B 2000 2016 2012 2074 In order to offer greater clarity to the reader,illustrates a perspective view of the fourth spectrometer. In this drawing, the device has been rotated so that it is primarily the second sideof the device (exposed) is facing toward the reader (rather than the opposing view from the first sidethat was shown in). In addition, for the sake of simplicity, the chassis protruding portion has been removed from sidewall.

21 FIG. 2184 2014 2184 In, a portion of laser diode moduleas mounted or connected to the device along the rearward sidecan be observed. In different embodiments, a laser printed circuit board can be incorporated into the apparatus that is capable of turning the laser on, off, and adjusting the power of the laser, allowing advanced users to fine-tune operational settings for their own purposes. In some embodiments, the apparatus can include an onboard computing device that can run a local AI model for onboard compound identification. The laser diode modulecan comprise any of the laser diode instruments, emitters, or arrays, described herein.

2000 2030 2030 2072 In different embodiments, the proposed fourth spectrometerincludes provisions for facilitating assembly of each module into the chassisas well as promoting the modular removal and replacement of these modules. For example, as noted earlier, chassisincludes multiple slotsof varying sizes, each slot sized and dimensioned to form a space of similar shape and size as a corresponding module.

2076 2004 2104 2006 2106 2052 2152 2054 2154 2156 2196 2158 2198 2058 2102 2060 2160 2160 2062 2162 21 FIG. More specifically, with respect to the components of optical routing portionshown in, the first lens moduleis slid into and received securely and snugly by a first slot, the second lens moduleis slid into and received securely and snugly by a second slot, the third lens moduleis slid into and received securely and snugly by a third slot, the beamsplitteris slid into and received securely and snugly by a fourth slot, the first filter componentis slid into and received securely and snugly by a fifth slotand the second filter componentis slid into and received securely and snugly by a sixth slot, the fourth lens moduleis slid into and received securely and snugly by a seventh slot, the pinhole moduleis slid into and received securely and snugly by an eighth slot(in this case, seventh slotis more of a narrow slit due to the narrow shape of the pinhole plate), and the achromatic lens moduleis slid into and received securely and snugly by a ninth slot.

2078 2000 2064 2066 2066 2176 2082 2176 2082 21 FIG. Furthermore, in different embodiments, the optical dispersive portionof fourth spectrometercan include provisions for securing and readily removing/replacing/adding individual metasurfaces from/to the device. In different embodiments, the optical compartmentincludes the dispersion unitwhich serves as a metasurface repository. In one embodiment, the dispersion unitcan include provisions by which to securely hold each metasurface in the correct position. For example, in, a plurality of grooved slots or channelsthat extend in a direction aligned with the longitudinal axiscan be seen. Each pair of channels are configured to receive and hold one metasurface. These channelscan be provided in pairs, so that for each metasurface there are two channels, each channel disposed along an opposing side, allowing each plate comprising a metasurface to be easily slid into place and securely oriented in a direction aligned with the longitudinal axis.

2062 2150 2176 2150 2150 2166 2064 21 FIG. 23 FIG.C Once installed, one face (the “leading surface”) of each metasurface can be oriented toward the achromatic lens module. in different embodiments, one or more metasurfacesare further secured in place from below, where the metasurface can drop into a groove formed along the base. Each groove can include a width that is sized and dimensioned to snugly receive the thickness of the metasurface. As shown in the example of, there are four pairs of channels, enabling the installation of four individual metasurfacesat the desired distance from one another. Thus, between each pair of metasurfaces, there is a gapto ensure the metasurfaces remained sufficiently spaced apart (e.g., see) from one another to allow the signal to continue to travel unimpeded through the optical compartmentfor optimal dispersion.

23 FIG.B 2124 2024 2064 2066 2060 2064 In different embodiments, the metasurfaces described herein are used to introduce controlled multiple scattering of the incoming Raman signal and generate unique speckle patterns. The multi-layered disordered metasurfaces (in this case, four) increase dispersion, enhancing spectral resolution across the near-infrared range (850-1100 nm). Each layer can be air-spaced at or around 1.5 mm interval gaps to optimize interference and improve wavelength discrimination (e.g., seebelow). A photodetector (not shown) can be mounted on a housing panelinstalled on the bottom side. The photodetector's sensor face can be directed into the chamber of the optical compartment, and receive the signal as it passes through the dispersion unitand exits the opening formed in the distal endof the optical compartment. In one embodiment, the metasurfaces transform spectral information into spatial speckle features that can be computationally reconstructed into a Raman signal. In some embodiments, the metasurfaces enable sub-nanometer resolution at the on-axis (along the same axis) photo detector (e.g., an IMX462). In some embodiments, one or more of the metasurfaces can include a disordered TiO2-coated metasurface. In one example, a metasurface can be sized with a 9×9 mm BK7 substrate (“plate”).

22 FIG. 2030 2000 Moving now to, additional portions of the chassishave been removed to expose and better illustrate the internal components/modules of the device. As noted earlier, the fourth spectrometercan include modular manufactured components and a corresponding chassis that form a sophisticated product. These pieces allow the spectrometer to operate using independent, interchangeable “modules” or blocks/units. These modules can be combined in various ways to create different configurations that meet specific needs, allowing for customization, flexibility, and efficiency.

22 FIG. 22 FIG. 2004 2202 2204 2006 2212 2210 2052 2206 2208 2058 2222 2224 2062 2226 2228 In, these modules (e.g., comprising a lens and a frame or slide) can be observed more clearly by the reader. In some embodiments, the modules can be configured as a block. For example, it can be seen inthat the first lens moduleis comprised of a first lensthat has been installed and secured in a first frame, second lens moduleis comprised of a second lensthat has been installed and secured in a second frame, third lens moduleis comprised of a third lensthat has been installed and secured in a third frame, and fourth lens moduleis comprised of a fourth lensthat has been installed and secured in a fourth frame, and the achromatic lens moduleis comprised of a fifth lensthat has been installed and secured in a fifth frame.

2030 2016 26 FIG. In different embodiments, each of the lens modules can take the form of a “lens block” that can be interchangeable with one another. In addition, each of the lens modules can be removed and replaced by simply sliding the block out of its associated slot and then inserting a new block into the same slot, so that any changes, swaps, repairs, etc. can occur in a few minutes. Once the housing has been removed and the sidewall of the chassisalong the top side, a swap can occur without any additional tools (e.g., replacement using hands/fingers only). Each block is robust and capable of withstanding repeated manual handling. In different embodiments, the device may be associated with or be part of a kit in which a variety of lens modules are included with differing specifications, size, and/or optical characteristics (e.g., collimating, chromatic/achromatic, etc.), allowing a user to readily modify the performance and output of the spectrometer as desired, update to a newer lens block model, or even simply remove a module if a particular block is undesirable. Additional details regarding these modular block units will be provided below with respect to.

2030 2000 2054 2030 2054 In different embodiments, other components installed in the chassisof fourth spectrometercan be similarly modular. For example, the beamsplittercomprises a substantially solid or continuous block that can be dropped into a corresponding slot provided by the chassis. The slot, comprised of a plurality of retaining ridges, can ensure the beamsplitteris mounted in its correct orientation and position relative to the surrounding lenses and filters.

2156 2214 2216 2158 2218 2220 In addition, the first filter component, which includes a first filtermounted and secured in a sixth frame, while second filter componentincludes a second filtermounted and secured in a seventh frame. The filter components can also be formed as blocks that can be readily removed and re-slotted and/or replaced by filters with different characteristics as needed by the user for a particular testing scenario.

2060 2260 2290 2030 2290 2030 2012 2060 Similarly, in different embodiments, the pinhole moduleincludes a plate or panel in which a pinhole is formed. The panel can be slotted and/or snugly installed into a grooved slit of a receiving blockattached/secured to a base portionof the chassis, where the base portioncorresponds to the sidewall of the chassisthat is provided along its second side. This pinhole modulecan also be readily replaced.

2100 Thus, while one assembly (such as assembly) can include the same types of components, by enabling layperson-friendly interchangeability of parts, the user can repeatedly, safely, and on-the-spot (e.g., “outside of a laboratory” or in the field) remove and replace one module with a first set of characteristics with another module with a second (different) set of characteristics. This approach allows the performance of the spectrometer to be dynamically fine-tuned or tweaked to better accommodate the properties of the sample being tested and/or specific requirements of a particular test.

2066 2150 2150 2066 2232 2234 2236 2238 2230 2230 2280 2290 2030 22 FIG. 24 25 25 FIGS.,A, andB As introduced above, in different embodiments, the dispersion unitcan also include provisions for modular interchangeability of the metasurfaces. In the drawing of, four metasurfacesare included in the dispersion unit, including a first metasurface, a second metasurface, a third metasurface, and a fourth metasurface. In some embodiments, each metasurface can be slotted into a corresponding groove formed in the thickness of a retaining portion. In addition, each groove can be aligned with and extend between a pair of channels, thereby forming a continuous three-sided slot for receiving and securing individual metasurfaces. In some embodiments, the retaining portioncan be mounted to a larger platform portionthat is attached/secured to the base portionof the chassis. Additional details regarding the metasurfaces will be provided with reference tobelow.

2030 2100 2024 2022 2100 23 23 23 FIGS.A,B, andC 23 FIG.A 23 FIG.B In order to better showcase the relative arrangement of the modules as installed in the chassis,depict a series of views of the assemblytaken from different perspectives. In, a direct bottom-side view is illustrated and in, a direct top-side view is illustrated, where each of the bottom sideand the top siderepresent opposing ends of the assembly.

23 FIG.A 2238 2230 2280 2280 2290 2238 2062 2062 2158 2052 2014 2238 2004 2010 2238 In the view of, the metasurfacesare shown mounted on retaining portion, which is disposed on the platform portion, where the platform portionis secured to base portion. In this embodiment, behind the metasurfacesis the achromatic lens module, and behind the achromatic lens moduleis the second filter component. Furthermore, the third lens moduleis visible, disposed along the rearward side(left of the metasurfaces), and the first lens moduleis visible, disposed along the forward side(right of the metasurfaces).

23 FIG.B 2006 2100 2054 2006 2052 2014 2006 2004 2010 2006 From the opposite end, shown in, the second lens modulecan be understood to provide the top-most component of the assembly(relative to the vertical axis). In this embodiment, the beamsplitteris visible, disposed behind the second lens module. In addition, the third lens moduleis visible, disposed along the rearward side(right of the second lens module), and the first lens moduleis visible, disposed along the forward side(left of the second lens module).

23 FIG.C 23 FIG.B 2006 2054 1 2054 2156 2 2 1 2 1 1 2 In, a side-view more clearly illustrates the relative positioning of each module and the spacings (identified as “distances” herein) between one module with its neighboring modules, with reference to the vertical axis. More specifically, from the most medial portion of the second lens module(associated with outermost point of the lens as it bulges medially away from its block housing) to a surface of the beamsplitterthere is a first distance D. In addition, between the beamsplitterand the first filter componentis a second distance D. In the embodiment shown in, Dis greater than D; however, in other embodiments, Dcan be substantially similar or equal to D, or Dcan be greater than D.

2156 2158 3 3 2 3 2 3 2 2158 2058 4 4 3 3 3 4 3 2058 2060 5 5 4 5 4 5 4 2060 2062 6 6 5 6 5 6 5 2062 2066 7 7 6 5 4 7 6 5 4 3 1 23 FIG.B 23 FIG.B 23 FIG.B 23 FIG.B 23 FIG.B Similarly, between the first filter componentand the second filter componentis a third distance D. In the embodiment shown in, Dis substantially similar or equal to D; however, in other embodiments, Dcan be greater than D, or Dcan be less than D. Furthermore, between the second filter componentand the fourth lens moduleis a fourth distance D. In the embodiment shown in, Dis greater than D; however, in other embodiments, Dcan be substantially similar or equal to D, or Dcan be less than D. Between the fourth lens moduleand the pinhole moduleis a fifth distance D. In the embodiment shown in, Dis greater than D; however, in other embodiments, Dcan be substantially similar or equal to D, or Dcan be less than D. In addition, between the pinhole moduleand the achromatic lens moduleis a sixth distance D. In the embodiment shown in, Dis greater than D; however, in other embodiments, Dcan be substantially similar or equal to D, or Dcan be less than D. Furthermore, between the achromatic lens moduleand the dispersion unitis a seventh distance D. In the embodiment shown in, Dis less than D, and also less than Dor D. However, in other embodiments, Dcan be substantially similar or equal to or less than D, D, D, D, and/or D.

2066 2150 2232 2234 8 2234 2236 9 2236 2238 10 8 9 10 8 9 9 10 8 10 23 FIG.B In addition, within the dispersion unititself, each metasurface of metasurfacescan be spaced apart from another neighboring or adjacent metasurface within the set. More specifically, the first metasurfaceand the second metasurfaceare spaced apart by an eighth distance D, the second metasurfaceand the third metasurfaceare spaced apart by a ninth distance D, and the third metasurfaceand the fourth metasurfaceare spaced apart by a tenth distance D. In the embodiment shown in, each of the distances D, D, and Dare substantially similar or equal. In other embodiments, these distances can differ from one another, so that for example, Dis greater than or less than D, Dis greater than or less than D, and/or Dis greater than or less than D.

In different embodiments, the distances between two metasurfaces and/or a metasurface and the achromatic lens module can be fine-tuned to produce a particular type of speckle pattern by the local device or to improve structural integrity between the components. For example, the distance between the achromatic lens module and the first metasurface can be approximately 2 mm, or at least 1.5 mm. In some embodiments, a minimum distance between each set of two neighboring metasurfaces can be approximately 1.5 mm. In addition, the photo detector (sensor) that will be disposed at the end of the optical compartment can be spaced apart from the final metasurface to minimize spectral interference. It can be appreciated that no additional lens is required to focus the signal onto the photo detector once the signal passes through the metasurfaces. In other words, in different embodiments, there are no additional components disposed between the last (most distal) metasurface disposed nearest to the sensor and the receptor of the sensor.

In different embodiments, the photo detector's receiving area can be sized as large as possible so there is an optimal use of the sensor for increasing or maximining resolution, and/or the number of metasurfaces can be greater, thereby increasing spectral resolution and accuracy. In some embodiments, compensation for intensity loss for spectral dissipation can include increasing exposure time (e.g., ten seconds versus five seconds). In one example, the proposed quadruple-layer metasurface assembly enables a spectral resolution over nearly every one of the wave numbers (e.g., over 80-90%), and in some cases to a thousandth of a wave number.

The proposed on-sensor configuration, consisting of the dispersive element (e.g., quadruple-layer disordered metasurfaces) accompanied by an on-axis image sensor, helps achieve robust spectrum reconstruction performance in a cost-effective manner. As one non-limiting example, in contrast to the capacity associated with conventional spectrometer (e.g., 1080×1920) which are limited to 1920 maximum points (because the grating breaks it up into a linear map), the metasurface approach described herein can take the 1920 points of measurability (i.e., those wavelengths and the intensity) and multiply it by 1080. This results in 2,073,600 points of reference, or three orders of magnitude (1000-fold) greater increase in information that can now be captured, even while continuing to measure the individual levels of intensity for each one of those pixels.

23 FIG.C 23 FIG.C 2100 2016 2100 2202 2204 1 2206 2208 2 2212 2210 3 2222 2224 4 2226 2228 5 The drawings ofalso allow the reader a better perspective from which to observe how, in different embodiments, the convex lens installed in each block protrudes or bulges outward relative to the periphery of the block. For example, in, which presents the side-view of the assemblyas well as a top-down view of the first sideof the assembly, the first lensprotrudes outward (medial direction) from its first frameby a distance P, the second lensprotrudes outward (medial direction) from its second frameby a distance P, the third lensprotrudes outward (medial direction) from its third frameby a distance P, the fourth lensprotrudes outward (top-wise direction) from its fourth frameby a distance P, and the fifth lensprotrudes outward (bottom-side direction) from its fifth frameby a distance P.

It can be appreciated that slots will be manufactured in the chassis that ensure the modules are inserted in the correct or desired spacing/distances between one another. In other words, each slot is carefully manufactured and placed to ensure the modules are oriented correctly when the user drops them in. It should further be appreciated that the distance by which a lens extends outward can contribute to its relative distance from a neighboring module, and in different embodiments, this will be accounted for when designing and manufacturing the slots in each location along the chassis. In some embodiments, this is because the slot geometry will generally be sized and dimensioned to receive a particular block size, regardless of whether the lens bulges outward from either side. In other words, when designing and manufacturing the chassis and its slot arrangement, each slot's relative position along the base portion and its distance to a neighboring slot can be adjusted to accommodate variations in the selected lens shapes and sizes, and more specifically to account for the extent or distance that a lens will bulge outward relative to the block in which it is secured, to help ensure the desired target optical characteristics/focal length, spectral resolution, and wavelength discrimination are achieved. Thus, in some embodiments, the chassis may be longer along the vertical axis to accommodate a greater distance between slots.

24 24 FIGS.A andB 24 FIG.A 2064 2064 offer further detail regarding the metasurfaces described herein. In, an isolated view of the optical compartmentis shown. In some embodiments, the optical compartmentcan include a plurality of substantially planar surfaces or sidewalls that collectively form a shape of a rectangular prism. In different embodiments, the interior of the compartment can be mostly or completely hollow and both the proximal end and the distal end include openings that allow clear, unobstructed access into and out of the hollow interior.

2064 2402 2404 2406 2402 2406 2404 2402 2406 2402 2066 2406 2064 2466 For reference, the optical compartmentcan include a first region, a second region, and a third region. When installed in the chassis, the first regionis most medial, the third regionthe most distal, while the second regionextends between the first regionand the third region. In this embodiment, the first regioncorresponds to the part or portion of the compartment where the dispersion unitis formed, while the distal most end of the third regioncan be where a sensor may be disposed/mounted (e.g., image sensor, CMOS sensor, or other sensor that can receive the signal that travels through the metasurfaces and from which speckle measurements can be obtained). The portion of the optical compartmentthat is most medial can be seen to include an opening or portthat allows light signals to travel from the achromatic lens module and pass into the interior of the compartment where the metasurfaces are installed.

2066 2066 2176 2064 24 FIG.A For purposes of illustration, the dispersion unitshown inincludes capacity for holding four metasurfaces. As noted above, the dispersion unitincludes pairs of channelsformed in at least two sidewalls of the optical compartment, as well as a groove running along the bottom. Each pair of channels and associated groove can receive one metasurface. In other words, a first peripheral edge of the metasurface can be slid into or snugly received by a first channel formed along a first sidewall, and a second peripheral edge (on the opposite side from the first peripheral edge) can be slid into or snugly received by a first channel formed along a second sidewall that is opposite to and faces the first sidewall. Furthermore, a third peripheral edge running orthogonally between the first peripheral edge and the second peripheral edge can be slid into or snugly received by a groove formed in the retaining portion below.

2234 2236 2238 2064 2432 2232 2434 2436 2438 2082 2230 2408 24 FIG.B In this drawing, second metasurface, third metasurface, and fourth metasurface(see) have been removed from the compartment to better reveal the interior configuration of the optical compartment. More specifically, in this example, there is a first pair of channels(currently holding the first metasurface), as well as a second pair of channelsconfigured to receive a metasurface, a third pair of channelsconfigured to receive metasurface, and a fourth pair of channelsconfigured to receive a metasurface. A set of parallel grooves (which can be oriented along the longitudinal axis) running across the width of the retaining portionat the bottom connect or join together each of the two channels in a pair to one another, and can be seen in this drawing (e.g., a first groove).

2066 2404 2402 2230 2404 2066 2066 2404 2406 2066 As noted earlier, in different embodiments, the number of metasurfaces can be changed and/or individual metasurfaces removed and/or replaced. In one embodiment, this can be accomplished simply by increasing or reducing the number of metasurfaces in the dispersion unit. However, in some cases additional metasurfaces (greater than four) may be desirable. In such instances, the optical compartment can be manufactured or modified so that some or all of the second regioncan be merged with the first region, and the retaining portionextended or elongated. Additional channels and grooves can be formed in this expanded section for receiving a greater number of metasurfaces. In other words, the second regioncan serve as a reserved area set aside in the case that more metasurfaces are to be added to the dispersion unit. In some embodiments, this expansion of the dispersion unitcan take up to the entire volume of the second region. However, it should be understood that in different embodiments, the third regionmay remain empty or cannot be adapted for metasurface installation/expansion of the dispersion unit, due to the need for a minimum spacing between the sensor and the most distally placed metasurface so that a signal can be captured by the sensor.

2232 7 2064 2466 2232 23 FIG.A Furthermore, in different embodiments, the distance between the first metasurfaceand the achromatic lens module can be larger than the seventh distance D(see), due to the additional spacing provided within the optical compartmentfrom its portto the first metasurface.

24 FIG.B 2238 2238 2430 2440 2430 2430 2410 2420 2430 2440 In order to better illustrate some of the features and performance characteristics associated with metasurfaces,presents an isolated view of the fourth metasurface. More specifically, the fourth metasurfaceincludes a medial sidethat receives incoming light signals from the lenses, and an opposing or opposite-facing distal sidefrom which the light signal is passed onward toward the sensor. To better appreciate the scale of the disordered structure formed on the medial side, a first section of the medial sideis shown at a first magnification level, and a smaller second section taken from within the first section is shown at a second magnification level, revealing an array of fabricated perturbations and disorder (“nanopost array”). It should be appreciated that in different embodiments, only one side (in this case, the medial side) includes these disordered physical characteristics (e.g., an array of nanoposts), while the opposite side (in this case, the distal side) can remain substantially smooth or lack such perturbations. In other words, only the very leading surface of a metasurface can include a disordered coating (e.g., of titanium dioxide), but not the back surface.

Within these magnified views, the disordered attributes of the metasurface become more apparent. In different embodiments, the disordered metasurfaces can serve as a spatio-spectral mixer, providing a versatile and complex mapping characteristic of high spectral sensitivity within a small footprint (e.g., 1 cm). By incorporating at least two of these metasurfaces into the optical system (thereby forming a “double-layer” or “stacked” arrangement of disordered metasurfaces), the system can provide two features: (a) predictability and (b) definitive mapping. The spectral response of the double-layer (or triple, quadruple, quintuple, and more, etc.) disordered metasurfaces may be seemingly random but they remain uniquely and accurately describable on the basis of the preconfigured design of disordered meta-atoms and a set of configurational parameters.

Sci. Adv. 2025 Additional details and features of the multi-layer disordered multisurfaces and reconstructive spectrometric techniques can be found in Dong-gu Lee et al., Reconstructive spectrometer using double-layer disordered metasurfaces.11, eadv2376() (DOI:10.1126/sciadv.adv2376) which is incorporated by reference herein in its entirety.

In different embodiments, using a multi-layered metasurface configuration to form a thin, forward-scattering medium for spectral-to-spatial mapping allows what might otherwise be a prohibitive calibration process to be relatively simplified on the basis of a chromato-axial memory effect where the same speckle pattern magnifies or demagnifies depending on the incoming wavelength. Thus, unlike conventional spectrometers that rely on a one-to-one mapping where each sensor element directly measures the intensity of a specific spectral band, the fourth spectrometer can incorporate a reconstructive approach involving a complex spectral-to-spatial mapping where the entire spectral information is sampled with a sensor array on a random basis and is subsequently decoded on the basis of the linear mapping relation.

In other words, the engineering domain of the disordered metasurface platform can be extended into the spatio-spectral domain, resolving the major challenges of conventional reconstructive spectrometers (e.g., the need for exhaustive calibration of output speckles for all independent spectral inputs). With the proposed arrangement of two or more metasurfaces, the characterization of absolute incoming wavelength and the reconstruction of continuous spectra can be achieved by feeding the output to an AI reconstruction module (e.g., running on an on-site or local mobile computing device) that can search/recognize a specific speckle pattern regardless of the constituent wavelengths of those individual speckle patterns. In different embodiments, because the disorder associated with each of the four (or more) metasurfaces will be random and different, there will be an AI-based customized calibration process performed for the individual spectrometer. The AI reconstruction module allows for the outputs from each individual spectrometer device—with its inherently different lens structure and metasurface dispersal—to be independently calibrated (i.e., AI-based local “self-calibration”). In one embodiment, an AI model of the AI reconstruction module can teach itself to recognize the speckle patterns being produced at the particular device. Furthermore, if at any point one or more modular components are replaced, one or more metasurface are removed, or additional metasurfaces added, an AI engine can automatically perform a recalibration to adapt to the impact on the speckle patterns and ensure they continue to be correctly recognized. This calibration can thereby be readily performed by a layperson, in the field, without any calibration experience.

In some embodiments, the calibration process can involve using a known pure sample of a certain chemical (such as but not limited to JP1 jet fuel). This pure sample (“reference sample”) can be provided with the device or obtained separately. The pure sample is then tested with the selected device and the collected pattern is calibrated with the identity of the chemical already known (e.g., assigning the speckle pattern generated as JP1 jet fuel). The AI engine can enter a calibration mode where it is understood that the data collected by the device at this time is part of a calibration step, and a pure sample is being used.

It can be appreciated that beyond the marked improvements in spectral resolution, this approach is robust against fabrication imperfections, as well as mechanical and thermal fluctuations. Unlike resonance-based approaches, the location of the correlation peak between the measured and computer-generated speckle maps (i.e., estimated wavelength) remains unaffected by fabrication errors in individual meta-atoms on the scale of tens of nanometers. In addition, the quadruple-layer (or more) configurations are characterizable with a handful of nondegenerate parameters, enabling seamless incorporation with computational optimization techniques. Unlike conventional spectrometers, which can be associated with ambiguity between free-space distance and incoming wavelength that necessitates frequent calibration of the one-to-one mapping relation between spectral and spatial domains to ensure diffraction-limited performance, the use of stacked disordered multi-layer metasurfaces can eliminate the need for frequent physical calibration of spectrometers that would require additional sources with known wavelengths. Because each device will have inherently different (unique) disordered patterns on each of metasurface in the stack of metasurfaces, the two-dimensional speckle patterns for a given chemical will be different from each other from one device to another, but will be predictable and consistent for each device, and the calibration (via local on-device AI self-teaching algorithms) will thereby be device-specific. No remote library of speckle patterns need be used as a reference, as the resultant speckle patterns and their meaning/chemical correspondence will be bound to the individual device itself.

In different embodiments, the AI reconstruction module can be fed with vast spectral libraries of biological and chemical signatures, and can be trained to transform raw photon scattering into plain-language identifications in seconds, without access to the cloud. While conventional spectrometer equipment typically output a cryptic spectrum for later analysis, the proposed spectrometer offers an instant verdict: for example, whether there is a pathogen present, at what concentration, and whether it is a lethal or non-lethal strain, etc.

25 25 FIGS.A andB 2078 2150 2236 2238 2076 2550 2150 2150 For clarity, additional details regarding the metasurfaces and the reconstructive approach are provided by reference to. In these two drawings, the optical dispersive portionthat includes the disordered metasurfacesand a CMOS sensor is presented in an exploded view. For convenience, only two of the metasurfaces (third metasurfaceand fourth metasurface) are shown here, but it should be understood by the ellipses ( . . . ) that additional metasurfaces can be disposed above (e.g., the first metasurface and second metasurface). In this example, the disordered double-layer metasurfaces serve as a random dispersive element that generates wavelength-specific speckle patterns. For example, as a sample is tested using the fourth spectrometer and its light signal is passed through the optical routing portionof the spectrometer, it can arrive at the optical compartment carrying an as-yet unknown spectrum. The signal can then travel through the random dispersive element corresponding to the spaced apart stack of metasurfaces. The metasurfacestransform the signal into wavelength-specific speckle patterns, which can be detected and measured by the on-axis sensor (e.g., disposed/installed at the most distal end of the optical compartment and being aligned with the center or main line of the metasurface's distal surface).

In different embodiments, the sensor can comprise of a monochrome photo detector that can offer higher quantum efficiency as well as eliminate ambiguous bare layer interaction within the individual pixels. A monochrome photo detector may be appropriate because the system is configured to look specifically for a particular speckle pattern, regardless of the constituent wavelengths of those individual speckle patterns. In other embodiments, a chromatic photo detector can be used.

In different embodiments, the measured intensity map can thus be represented as the superposition of speckle patterns on the basis of independent spectral channels. For example, in some embodiments, the design of the disordered metasurfaces and the configurational parameters are used to construct a local-device-specific based computer-generated speckle library. In such cases, this custom library cannot be generalized for use by other devices, as it will be based on calibration that occurs at the local device. In some embodiments, this local library can then be subsequently used to reconstruct the spectrum by solving the inverse problem represented as matrix-vector multiplication, i=Ws, where i is the vectorized measured intensity, W is spectral response matrix, and s is the input spectrum.

x 2 3 2 3 x x 4 8 6 2 3 4 2 2 In some embodiments, a disordered metasurface can be fabricated with silicon nitride (SiN) or other similar material that is used to form nanoposts of randomized widths across one surface, with associated phase delay values ranging from 0 to 2π at the design wavelength (e.g., 532 nm). As a non-limiting example, fabrication can begin where, on one side of the substrate, a pattern is formed on a fused silica substrate by transferring the pattern using photolithography and lift-off techniques (e.g., with layers of 10 nm Cr and 100 nm Au). A 60-nm-thick aluminum oxide (AlO) layer is deposited onto a fused silica substrate through e-beam evaporation. The resist can then be stripped using resist remover. The resulting AlOpattern on SiNcan be transformed into a SiNnanopost array by inductively coupled plasma reactive ion etching (ICP-RIE) using a mixture of CFand SFgases. The residual AlOmask can be removed using a mixture of NHOH and HO. In some embodiments, to configure multi-layer metasurfaces, the space between each metasurface substrate and the next can be filled with 3D-printed spacer molds and fixed using ultraviolet-cured resin. In other embodiments, the channels formed in the sidewalls and grooves formed in the retaining portion for receiving each metasurface can be sufficient to hold and fix the metasurfaces in the target configuration.

25 FIG.B When stacking two metasurfaces in front of an image sensor, there are two major variables, T and L, the thickness of the random dispersive element composed of the two metasurfaces and the element-to-sensor distance, respectively. Those two variables determine the spectral resolution, δλ, and the sampling condition for speckles in conjunction with the additional variables such as the aperture size of disordered metasurfaces, D, and the pixel size of the image sensor, Δp. A disordered metasurface can be considered as a random spatial mixer that simultaneously generates many plane wave components within a confined spatial frequency range. Because of the angular dispersion effect, the interfering plane waves with random phases generate a wavelength-dependent speckle pattern when given sufficient propagation distance. Then, the spectral resolution, δλ, can be defined as the full-width at half maximum (FWHM) of the spectral correlation profile for the generated speckle patterns. In different embodiments, the speckle output at a specific wavelength and configuration can then be predicted through a parameterized wave propagation model, as shown in.

26 27 28 29 FIGS.,,, and 26 FIG. 26 FIG. 2610 2620 Additional information regarding the use of modular blocks for the spectrometer will be now be presented with reference to. In, an exploded view of an embodiment of a plurality of framesthat can be installed in a chassis is depicted, alongside an exploded view of an embodiment of a plurality of lensesthat can be mounted in each frame. In, the modularity of the block units can be better appreciated, here comprising discrete two-piece components (i.e., where the lens is mounted within and retained by a protective frame).

26 FIG. 2602 2208 2602 2612 2614 2604 2604 2602 For example, in, an interior surface of a tunnelformed in the second framecan be observed. In different embodiments, the tunnelis sized and dimensioned to snugly receive a lens. In some embodiments, the interior can include a lens plano body sleeve portionand a convex lens retaining portion. Between the two portions can be a retaining rimthat extends around the circumference of the interior of the tunnel. The retaining rimcan slightly reduce the size of the tunnel by providing a ridge that extends radially-inward, and is configured to prevent the plano body portion of the lens from moving further out of the tunnelonce it is slide into the tunnel, while at the same time permitting the convex portion to bulge outward from the other side of the tunnel.

2208 2206 2622 2624 2632 2602 2022 2086 2024 2632 2624 2622 2604 For example, the corresponding lens for the second frame, shown here as the second lens, can include a plano body portionand a bulging convex portionjoined along an intermediate zone. When inserted into the tunnelfrom the end that is facing top side, the entire lens can move or be pushed in a direction aligned with vertical axis(toward bottom side) until the intermediate zone(signaling the transition point between the convex portionand the plano body portion) contacts the retaining rim, which stops further movement in that direction and holds the lens in place.

27 FIG. 2700 2010 2124 2790 2780 For clarity,presents an example of an assemblyin which only the lenses are shown in their associated slots that are formed in the chassis. In other words, for each module, the outermost frame has been removed to reveal the lens that is mounted within. In different embodiments, these components are formed only of glass, as represented by the series of lines drawn across the surfaces of the components. In other embodiments, the components can include another monolithic or polymer optical lens material. In this view, the housing panelto which an embodiment of a photodetectoris mounted can also be observed, as well as a portion of a laser diode receiving compartment.

28 FIG. 2700 2710 2720 2710 2710 2720 Thus, as shown in, a first groupof block units can be installed in a chassis, where each block unit comprises of an outer frameand a lensthat has been mounted inside of the frame. The outer frameand the lenscan be comprised of two different materials (e.g., glass in the lens and non-glass in the frame).

29 FIG. 28 FIG. 2900 2800 It can be appreciated that in some embodiments, the manufacture of a two-piece unit may be more complex than that of a one-piece unit. In some embodiments, the spectrometer can alternatively include modular units that are formed as monolithic objects. In other words, the entire module can be manufactured as a single, continuous, unbroken object comprising only one type of material. This material can be the same material that is used in making the desired lens (e.g., glass). In such cases, as shown in, the assembly for a modular spectrometer can include a chassis as described herein but with a different, second groupof module blocks. In different embodiments, these modules can be installed in the same chassis and offer the same functionality as the first groupof, with significantly less cost and complexity in their manufacture. Furthermore, the precision associated with each unit can be increased as there is no longer a mechanical insertion process of the lens into the frame, the retaining rim will no longer be required, and the size and dimensions of the convex portion relative to the plano portion and where they meet can be built with much more exacting specification.

2810 2904 2906 2952 2958 2964 2956 2954 2700 For example, in different embodiments, each full or complete block unit/module can instead be made of a single, solid, seamless piece, comprising the same material (such as, but not limited to, glass or polycarbonate). In different embodiments, each of a first monolithic lens module, second monolithic lens module, third monolithic lens module, fourth monolithic lens module, monolithic achromatic lens module, first monolithic filter component, and second monolithic filter componentcan be used interchangeably with the blocks of the first groupin the same chassis. In one embodiment, the assembly of the device can include a step of simply dropping, inserting, or sliding each of these monolithic modules into their associated respective slot formed in the chassis as a set of whole, solid, continuous, and/or one-piece units.

30 FIG. 3010 3020 3030 Turning to, in order to offer a broader context to the reader, several use-case scenarios employing an embodiment of the spectrometer described herein are depicted. A first exampleshows how the spectrometer, with an embedded AI system, is both lightweight and small enough to be mounted on and transported by medium or large (e.g., 5-7 inches) drones that can then travel to dangerous areas to perform testing or other environmental monitoring. A second exampleshows how the spectrometer can serve as wall-mounted sentinels, such as in hospitals and transportation hubs, for continuous monitoring of airborne pathogens, enabling detection of outbreaks before symptomatic cases present. Such a sentinel can be compact and easy to mount and/or remove from the wall. A third exampledepicts a spectrometer that has been placed in a sewer for bio-surveillance. In still other examples, the spectrometer can be integrated into water-testing stations for continuous monitoring of pathogens and toxins in drinking water, and deep-sea gas monitoring. In another example, the spectrometer may be handheld and carried by health workers for rapid pathogen identification at point-of-care and/or rapid differentiation of bacterial strains. In some cases, the spectrometer may be handheld and carried by agriculturists or food health and safety personnel to perform non-destructive detection of any contaminants and pesticide residues.

31 FIG. 3100 3100 3110 3120 3130 3140 3150 is a flow chart illustrating an embodiment of a methodof detecting contaminants in a chemical sample. The methodincludes a first stepof sending, from a chemical analysis application (“app”) installed on a mobile computing device, a control signal to a handheld spectrometer apparatus connected to the mobile computing device that causes the spectrometer apparatus to perform a first test cycle involving a first sample. A second stepincludes receiving, at the app and from the spectrometer apparatus, first image data captured by a photoreceptor of the spectrometer apparatus, the first image data including spectral data for the first sample. In a third step, the method includes passing the first image data to a deep neural network (DNN) model that is trained to detect and quantify, in spectral data, one or more contaminants of a plurality of potential contaminants. In different embodiments, these potential contaminants can include Polyalphaolefin (PAO), sulfur compound(s), synthetic fuel additive(s), hydraulic fluid(s), and microbial compound(s). A fourth stepincludes determining, via the DNN model and based on the first image data, the first sample includes a first contaminant, and a fifth stepincludes presenting, via a graphical user interface (GUI) for the app, a notification indicating the first sample includes the first contaminant.

3100 In different embodiments, the methodmay include additional steps or aspects. In some embodiments, the method also includes receiving, at the app, a first input from a user selecting a first sample profile, wherein the control signal is sent in response to receiving the first input. In another embodiment, the method also includes presenting, via the GUI, a plurality of selectable options, each selectable option identifying a different sample profile, where the first input corresponds to a selection of one of the plurality of selectable options. In some embodiments, determining the first sample includes the first contaminant is based on the user selection of the first sample profile. In one embodiment, the method further includes receiving, at the app, a first input from a user selecting a first sample profile, where selection of a sample profile is used by the app to limit detection to a subset (less than the total that can be tested by the system) of the plurality of potential contaminants. In other words, in some embodiments, the pre-configured profile that is selected by the user defines which contaminants should be flagged if detected in the sample. In another embodiment, the method includes presenting, via the GUI, a spectral graph plotting an intensity of scattered light versus a frequency of light as characterized by the first image data.

As described herein, some of the proposed embodiments can also be understood to include a man-portable apparatus for identification of compounds. The apparatus can include: (a) a housing including a first compartment and a second compartment; (b) a light source that directs light into the first compartment; and (c) an optical system including a dichroic beamsplitter and a collection lens, wherein the collection lens is oriented along a first axis, and the dichroic beamsplitter is oriented along a second axis that is at a 45-degree angle relative to the first axis.

In other embodiments, the apparatus may include additional features, components, or aspects. In some embodiments, the housing further includes a slit portion disposed between the first compartment and the second compartment, and light reflected from the dichroic beamsplitter in the first compartment passes through the slit portion and into the second compartment. In another embodiment, the light source includes a laser diode oriented along a third axis that is at a 45-degree angle relative to the second axis. In different embodiments, the apparatus further includes a grating in the second compartment oriented at an acute angle relative to the third axis. In one embodiment, the apparatus also includes a Raleigh filter installed along a sidewall of the second compartment, where light reflected from the grating passes through the Raleigh filter. In some embodiments, the apparatus further includes a photoreceptor mounted on an exterior of the sidewall of the second compartment adjacent to the Raleigh filter, and the light exiting the Raleigh filter is captured by the photoreceptor. In different embodiments, the apparatus also includes a computer processor that is configured to share image data captured by the photodetector to a mobile computing device. In some embodiments, the apparatus includes a connector element protruding from an exterior of the housing, the connector element being configured to connect the apparatus to a data port of a mobile computing device.

In different embodiments, some of the proposed embodiments can also be understood to include a kit of parts for performing chemical analyses. The kit can include a handheld spectrometer apparatus and a mobile computing device. The spectrometer apparatus can include an optical system comprising a dichroic beamsplitter and a collection lens. In some embodiments, the collection lens is oriented along a first axis, and the dichroic beamsplitter is oriented along a second axis that is at a 45-degree angle relative to the first axis. The spectrometer apparatus can also include a male connector element. In addition, the mobile computing device can include a female data port that is configured to connect to the male connector element and enable communication between the spectrometer apparatus and the mobile computing device.

In other embodiments, the kit of parts may include additional features, components, or aspects. In some embodiments, the kit also includes a chemical analysis software application (“app”) installed on the mobile computing device, the app being configured to receive spectral data for a sample from the spectrometer apparatus and predict chemical properties of the sample based on the spectral data using an artificial intelligence (AI) neural network. In one embodiment, the spectrometer apparatus includes a photoreceptor, and the spectral data is conveyed by images captured by the photoreceptor. In some embodiments, the app includes a Deep Neural Network (DNN) model that is trained to detect contaminants in a test sample based on the image data. In different embodiments, the kit also includes a vial for holding a sample to be tested, and the spectrometer apparatus further includes a cylindrical receptacle that is sized and dimensioned to snugly receive the vial. In some embodiments, the cylindrical receptacle is adjacent to the collection lens, and a center of the vial is disposed at a distance of between 15 mm and 25 mm when the vial is inserted into the receptacle. In different embodiments, the kit can also include a microfluidic slide for holding a sample to be tested, and the spectrometer apparatus further includes a slot that is sized and dimensioned to snugly receive the slide. In some embodiments, the spectrometer apparatus receives power from the mobile computing device when the spectrometer apparatus is connected to the mobile computing device.

In different embodiments, the kit also includes a first modular sample holder that can be removably attached to a first portion of a housing of the spectrometer apparatus adjacent to the collection lens, the first modular sample holder including a first receptacle for receiving a first vial of a first size, and a second modular sample holder that can be removably attached to the first portion of the housing and includes a second receptacle for receiving a second vial of a second size, the first size being smaller than the second size. In other embodiments, the kit includes a first modular sample holder that can be removably attached to a first portion of a housing of the spectrometer apparatus adjacent to the collection lens, the first modular sample holder including a first cylindrical receptacle for receiving a vial; and a second modular sample holder that can be removably attached to the first portion of the housing and includes a second slotted receptacle for receiving a microfluidic slide.

In different embodiments, some of the proposed embodiments can also be understood to include a microfluidic chip for use with a handheld spectrometer apparatus. The microfluidic chip includes (a) a substantially planar substrate; (b) an optical window integrated into the substrate; (c) a filter embedded in a center of the optical window; (d) a sample port formed at a lower end of the substrate; and (e) a plurality of reaction wells etched or molded into the substrate, each reaction well extending independently from the optical window before merging and connecting to the sample port.

In other embodiments, the microfluidic chip may include additional features, components, or aspects. In some embodiments, the optical window is formed with quartz and is configured to enhance a Raman signal for clarity in trace chemical detection by the spectrometer apparatus. In different embodiments, one or more of the reaction wells of the plurality of reaction wells includes colorimetric compounds. In some embodiments, the microfluidic slide causes preprocessing chemical interactions in a sample that enable the spectrometer apparatus to test a sample for one or more of total acid number, viscosity, density, cetane index, distillation fractions, freezing point, flash point, smoke point, cloud point, neat heat of combustion, MSEP rating, copper strip corrosion, thermal oxidation stability, electrical conductivity, existent gum content, fuel lubricity, and/or carbon residue. In different embodiments, the optical window is disposed in a central region of the substrate so that the optical window aligns with a focal point of the spectrometer apparatus when the microfluidic chip is inserted into the spectrometer apparatus.

32 FIG. 3200 3210 3200 3220 3230 3240 3250 Other methods may be contemplated within the scope of the present disclosure. For example, referring to, in some embodiments, a methodof refining a global neural network model for performing chemical analyses of spectral data is provided. The method can include a first stepof receiving, at a federated server, first data from a first local model running on a first mobile computing device at a first location, the first data generated based on spectral data captured by a first spectrometer apparatus connected to the first mobile computing device. The methodcan also include a second stepof receiving, at the federated server, second data from a second local model running on a second mobile computing device at a second location, the second data generated based on spectral data captured by a second spectrometer apparatus connected to the second mobile computing device. A third stepincludes training, at the federated server, the global neural network model using an aggregation of the first data and the second data. In addition, a fourth stepincludes transmitting, from the federated server and to the first mobile computing device, an updated set of weights for implementation by the first local model. Furthermore, a fifth stepincludes transmitting, from the federated server and to the second mobile computing device, the updated set of weights for implementation by the second local model.

In other embodiments, the method may include additional steps or aspects. In some embodiments, the first data includes one or more of gradients, weights, and performance metrics from the first local model. In one embodiment, the first data is anonymized and encrypted at the first local model before transmission to the federated server. In different embodiments, the global neural network model is a deep neural network (DNN) that applies Continuous Wavelet Transformation (CWT) to detect and quantify one or more of Polyalphaolefin (PAO), sulfur compounds, synthetic fuel additives, hydraulic fluid, and microbial contamination in a sample. In some embodiments, the method further includes training the global neural network model using synthesized training data. In different embodiments, generation of the synthesized training data involves converting mass spectrometry (MS) data to Raman spectrometry data. In some embodiments, the synthesized training data is generated via an automated Density Functional Theory (DFT) algorithm that can create artificial Raman spectra. In one embodiment, the global neural network model includes an anomaly detection algorithm for identifying outlier patterns indicating contaminants in jet fuels.

Other methods may be contemplated within the scope of the present disclosure. For example, in some embodiments, a method performing spectral collection and analysis in a field setting is provided. The method can include a first step of connecting a handheld spectrometer apparatus to a mobile computing device, and a second step of connecting a first modular sample holder to the spectrometer apparatus. A third step includes inserting a first vial containing a first sample into an aperture formed in the first modular sample holder, the aperture sized and dimensioned to snugly receive the first vial, and the aperture being disposed adjacent to a collection lens of the spectrometer apparatus. In addition, a fourth step can include initiating, via a chemical analysis software application (“app”) installed on the mobile computing device, a first test cycle, thereby causing the spectrometer apparatus to pass light from a laser diode to a dichroic beamsplitter with a face oriented at a 45-degree excitation angle relative to both the laser diode and the collection lens. Furthermore, a fifth step can include receiving, via a graphical user interface (GUI) provided by the app, a chemical and structural characterization of the first sample.

In other embodiments, the method may include additional steps or aspects. In some embodiments, the method can include: disconnecting the first modular sample holder from the spectrometer apparatus; connecting a second modular sample holder to the spectrometer apparatus; and inserting a first microfluidic chip containing a second sample into a slot formed in the second modular sample holder, the slot sized and dimensioned to snugly receive the first microfluidic chip, and the slot being disposed adjacent to a collection lens of the spectrometer apparatus. In some embodiments, the method can include steps of initiating, via the app, a second test cycle, thereby causing the spectrometer apparatus to pass light from a laser diode to a dichroic beamsplitter with a face oriented at a 45-degree excitation angle relative to both the laser diode and the collection lens; and receiving, via the GUI, a chemical and structural characterization of the second sample.

In different embodiments, the method can include inserting, into a housing of the spectrometer apparatus, a first transmission grating with a first wavelength range for use during the first test cycle. In some embodiments, the method also includes removing the first transmission grating from the spectrometer apparatus; and inserting, into the spectrometer apparatus, a second transmission grating with a second wavelength range for use during the second test cycle, the second wavelength range differing from the first wavelength range. In some embodiments, the method can include selecting, via the app and before the first test cycle, a first fuel profile that targets detection of a first contaminant, wherein any contaminants identified in the chemical and structural characterization of the first sample are based on the selection of the first fuel profile.

In different embodiments, the method can include connecting, to a housing of the spectrometer apparatus, a first photoreceptor with a first quantum efficiency for use during the first test cycle. In some embodiments, the method can further include steps of: removing the first photoreceptor from the spectrometer apparatus; and connecting, to the spectrometer apparatus, a second photoreceptor with a second quantum efficiency for use during the second test cycle, the second quantum efficiency differing from the first quantum efficiency.

In different embodiments, some of the proposed embodiments can also be understood to include a man-portable spectrometer system. The man-portable spectrometer system includes (a) an optical routing portion including a first lens; (b) an optical dispersive portion including a dispersion unit, wherein the dispersion unit is configured to receive an incoming light signal associated with a sample that is routed from the optical routing portion to the optical dispersive portion, and includes a plurality of disordered metasurfaces; (c) a photo detector configured to receive an output of the light signal as a wavelength-dependent speckle pattern as the light signal exits the optical dispersive portion; and (d) an artificial intelligence (AI) spectrum reconstruction model configured to receive the speckle pattern and automatically recognize one or more chemical compounds in the sample based on the speckle pattern.

In other embodiments, the spectrometer system may include additional features, components, or aspects. In some embodiments, the plurality of disordered metasurfaces includes a first metasurface with a proximal surface oriented toward the optical dispersive portion and an opposite-facing distal surface oriented toward the photo detector. In one embodiment, the proximal surface includes a nanopost array of randomized widths that serve as a random spatial mixer for the incoming light signal. In one embodiment, the distal surface of the first metasurface is substantially smooth.

In different embodiments, the optical routing portion includes a first lens that is oriented toward the proximal surface of the first metasurface. In some embodiments, the optical routing portion includes a first filter that is oriented toward the proximal surface of the first metasurface. In one embodiment, the plurality of disordered metasurfaces includes at least three metasurfaces that are on-axis with respect to one another. In some embodiments, the photo detector is on-axis with respect to each metasurface of the plurality of disordered metasurfaces. In different embodiments, the dispersion unit is housed within an optical compartment, the optical compartment includes a first sidewall, and the first sidewall includes a plurality of channels, wherein each channel of the plurality of channels is sized and dimensioned to snugly receive a peripheral edge of one metasurface of the plurality of disordered metasurfaces.

In different embodiments, some of the proposed embodiments can also be understood to include a modular assembly for a spectrometer. The assembly includes (a) a chassis including a plurality of slots, the plurality of slots including a first slot and a second slot; (b) a plurality of interchangeable modules including a first module and a second module, where the first slot is sized and dimensioned to snugly receive the first module and the second slot is sized and dimensioned to snugly receive the second module; and (c) where the first module is one of a lens, an optical compartment, a beamsplitter, a pinhole panel, and a filter.

In other embodiments, the spectrometer system may include additional features, components, or aspects. In some embodiments, the first module and second module are interchangeable with one another, such that the first slot is also sized and dimensioned to snugly receive the second module and the second slot is sized and dimensioned to snugly receive the first module. In another embodiment, the first module includes a lens mounted within an opening formed in a block. In different embodiments, the first module is a monolithic object comprising a glass lens formed within (as part of/integral with) a glass block.

In some other embodiments, the second module is a pinhole panel and the second slot comprises a groove formed within two sides of the chassis that receives two opposing edges of the pinhole panel. In some embodiments, the second slot further comprises a slit formed in a retaining block secured to a base of the chassis, and the slit receives a third edge of the pinhole panel running orthogonal to the two opposing edges. In different embodiments, the first module is the optical compartment, and the optical compartment includes an interior space configured to hold a plurality of disordered metasurfaces. In some embodiments, the plurality of disordered metasurfaces include a first metasurface and a second metasurface, and the interior space includes a first pair of channels and a second pair of channels, and the first pair of channels is sized and dimensioned to snugly receive the first metasurface and the second pair of channels is sized and dimensioned to snugly receive the second metasurface. In different embodiments, the first metasurface and second metasurface are interchangeable with one another, such that the first pair of channels is also sized and dimensioned to snugly receive the second metasurface and the second pair of channels is sized and dimensioned to snugly receive the first metasurface.

In some embodiments, the first module includes a filter mounted within an opening formed in a block. In another embodiment, the first module is a monolithic object comprising a filter formed within a block, and the filter and block are made of the same material

Other methods may be contemplated within the scope of the present disclosure. For example, in some embodiments, a method of performing on-device calibrations of a spectrometer device that includes a plurality of disordered metasurfaces is provided. The method can include a first step of testing a reference sample using the spectrometer device, wherein the reference sample consists of a pre-identified first chemical; a second step of capturing a first speckle pattern via a photo detector of the spectrometer device; and a third step of characterizing, by a local artificial intelligence (AI) spectrum reconstruction engine, the first speckle pattern as representative of the first chemical.

In other embodiments, the method may include additional steps or aspects. In some embodiments, the method can include switching operations of the spectrometer device to its calibration mode before testing the reference sample occurs. In different embodiments, testing of the reference sample further includes operations of: receiving, at an optical routing portion of the spectrometer device, a light signal emitted from the reference sample; and transmitting the light signal from the optical routing portion to an optical dispersive portion that includes the plurality of disordered metasurfaces. In some embodiments, the first speckle pattern is generated as a result of the light signal passing through the plurality of disordered metasurfaces.

In different embodiments, any changes to the modules or elements thereof can require or necessitate a re-calibration session. For example, in different embodiments, where the plurality of disordered metasurfaces is ordered in a first configuration, the method can further include rearranging the plurality of disordered metasurfaces to a different, second configuration; and then (in response to the rearrangement) performing another calibration session by: retesting the reference sample using the spectrometer device; capturing a second speckle pattern via the photo detector; and characterizing, by the local AI spectrum reconstruction engine, the second speckle pattern as representative of the first chemical.

In some embodiments, the plurality of disordered metasurfaces includes a first metasurface, and the method further includes: replacing the first metasurface with a different, second metasurface; and then (in response to replacing any of the metasurfaces with a different metasurface) performing another calibration session by: retesting the reference sample using the spectrometer device; capturing a second speckle pattern via the photo detector; and characterizing, by the local AI spectrum reconstruction engine, the second speckle pattern as representative of the first chemical.

In another embodiment, the plurality of disordered metasurfaces includes three or more metasurfaces, and the method further includes: removing one metasurface from the three or more metasurfaces; and then (in response to removing a metasurface) performing another calibration session by: retesting the reference sample using the spectrometer device; capturing a second speckle pattern via the photo detector; and characterizing, by the local AI spectrum reconstruction engine, the second speckle pattern as representative of the first chemical.

In some embodiments, the method can also include: increasing the number of metasurfaces of the spectrometer device; and then (in response to adding one or more additional metasurfaces) performing another calibration session by: retesting the reference sample using the spectrometer device; capturing a second speckle pattern via the photo detector; and characterizing, by the local AI spectrum reconstruction engine, the second speckle pattern as representative of the first chemical. Thus, with each change, the output from the dispersion unit will shift, and local calibration must be performed again for the device to ensure the test results continue to be correctly characterized by the AI model.

As described herein, the proposed systems, kits, methods, and apparatuses are configured to offer optimized real-time analysis capabilities through advanced signal processing techniques, including multivariate chemometric modeling and machine learning algorithms. The implementation of sensor fusion in the system provides a complementary approach, where the use of GC enables hydrocarbon separation and quantification, and Raman spectroscopy enables rapid molecular identification. By integrating these techniques with CWT signal processing and AI-based analytical techniques, the systems will mitigate environmental noise, improve detection limits, and facilitate real-time American Society for Testing and Materials (ASTM) International-compliant fuel analysis. By manifesting the system as a compact, field-deployable system capable of real-time fuel and propellant analysis in challenging operational environments, laymen will be able to perform rapid, accurate assessments of critical chemical properties, including sulfur content, cetane number, water contamination, and molecular composition. In cases where aircraft and other fuel-propelled vehicles are to be monitored, this field portable device can significantly enhance mission readiness by eliminating the delays associated with laboratory-based testing, which often require extensive logistical support, or other conventional testing equipment which are bulky, complex, and limited in their analytical capabilities.

Furthermore, in different embodiments, the proposed devices can incorporate a modular system for performing spectroscopy, including interchangeable “blocks” that can be dropped into corresponding slots formed in the housing, and in some cases using monolithic lens-only “glass blocks” (no outer casing) for insertion into a device, helping to overcome significant manufacturing challenges that may be otherwise associated with the production of the device.

By enabling ASTM-compliant, real-time testing in the field, the system will mitigate risks associated with fuel contamination and quality failures, enhancing both safety and operational success. The device will also yield significant cost savings by reducing reliance on centralized laboratory testing and streamlining logistics. Expanded capabilities will also include measurements such as total acid number, viscosity, density, cetane index, distillation fractions, freezing point, flash point, smoke point, cloud point, neat heat of combustion, MSEP rating, copper strip corrosion, thermal oxidation stability, electrical conductivity, existent gum content, fuel lubricity, and carbon residue. These expansions will build on the same core principles while integrating new technologies to enhance analytical versatility.

It is to be appreciated that methods in accordance with the present disclosure can include any combination of the aspects and features described herein. That is, methods and systems in accordance with the present disclosure are not limited to the combinations of aspects and features specifically described herein, but also include any combination of the aspects and features provided.

Throughout this application, an “interface” may be understood to refer to a mechanism for communicating content through a client application to an application user. In some examples, interfaces may include pop-up windows that may be presented to a user via native application user interfaces (UIs), controls, actuatable interfaces, interactive buttons or other objects that may be shown to a user through native application UIs, as well as mechanisms that are native to a particular application for presenting associated content with those native controls. In addition, the terms “actuation” or “actuation event” refers to an event (or specific sequence of events) associated with a particular input or use of an application via an interface, which can trigger a change in the display of the application. This can include selections or other user interactions with the application, such as a selection of an option offered via a native control, or a ‘click’, toggle, voice command, or other input actions (such as a mouse left-button or right-button click, a touchscreen tap, a selection of data, or other input types). Furthermore, a “native control” refers to a mechanism for communicating content through a client application to an application user. For example, native controls may include actuatable or selectable options or “buttons” that may be presented to a user via native application UIs, touch-screen access points, menus items, or other objects that may be shown to a user through native application UIs, segments of a larger interface, as well as mechanisms that are native to a particular application for presenting associated content with those native controls. The term “asset” refers to content that may be presented in association with a native control in a native application. As some non-limiting examples, an asset may include text in an actuatable pop-up window, audio associated with the interactive click of a button or other native application object, video associated with a teaching/tutorial user interface, or other such information presentation.

The processes and methods of the embodiments described in this detailed description and shown in the figures can be implemented using any kind of computing system having one or more central processing units (CPUs) and/or graphics processing units (GPUs). The processes and methods of the embodiments could also be implemented using special purpose circuitry such as an application specific integrated circuit (ASIC). The processes and methods of the embodiments may also be implemented on computing systems including read only memory (ROM) and/or random access memory (RAM), which may be connected to one or more processing units. Examples of computing systems and devices include, but are not limited to: servers, cellular phones, smart phones, tablet computers, notebook computers, e-book readers, laptop or desktop computers, all-in-one computers, as well as various kinds of digital media players.

The processes and methods of the embodiments can be stored as instructions and/or data on non-transitory computer-readable media. The non-transitory computer readable medium may include any suitable computer readable medium, such as a memory, such as RAM, ROM, flash memory, or any other type of memory known in the art. In some embodiments, the non-transitory computer readable medium may include, for example, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of such devices. More specific examples of the non-transitory computer readable medium may include a portable computer diskette, a floppy disk, a hard disk, magnetic disks or tapes, a read-only memory (ROM), a random access memory (RAM), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), an erasable programmable read-only memory (EPROM or Flash memory), electrically erasable programmable read-only memories (EEPROM), a digital versatile disk (DVD and DVD-ROM), a memory stick, other kinds of solid state drives, and any suitable combination of these exemplary media. A non-transitory computer readable medium, as used herein, is not to be construed as being transitory signals, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

Instructions stored on the non-transitory computer readable medium for carrying out operations of the present invention may be instruction-set-architecture (ISA) instructions, assembler instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, configuration data for integrated circuitry, state-setting data, or source code or object code written in any of one or more programming languages, including an object oriented programming language such as Smalltalk, C++, or suitable language, and procedural programming languages, such as the “C” programming language or similar programming languages.

Aspects of the present disclosure are described in association with figures illustrating flowcharts and/or block diagrams of methods, apparatus (systems), and computing products. It will be understood that each block of the flowcharts and/or block diagrams can be implemented by computer readable instructions. The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of various disclosed embodiments. Accordingly, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions. In some implementations, the functions set forth in the figures and claims may occur in an alternative order than listed and/or illustrated.

The embodiments may utilize any kind of network for communication between separate computing systems. A network can comprise any combination of local area networks (LANs) and/or wide area networks (WANs), using both wired and wireless communication systems. A network may use various known communications technologies and/or protocols. Communication technologies can include, but are not limited to: Ethernet, 802.11, worldwide interoperability for microwave access (WiMAX), mobile broadband (such as CDMA, and LTE), digital subscriber line (DSL), cable internet access, satellite broadband, wireless ISP, fiber optic internet, as well as other wired and wireless technologies. Networking protocols used on a network may include transmission control protocol/Internet protocol (TCP/IP), multiprotocol label switching (MPLS), User Datagram Protocol (UDP), hypertext transport protocol (HTTP), hypertext transport protocol secure (HTTPS) and file transfer protocol (FTP) as well as other protocols.

Data exchanged over a network may be represented using technologies and/or formats including hypertext markup language (HTML), extensible markup language (XML), Atom, JavaScript Object Notation (JSON), YAML, as well as other data exchange formats. In addition, information transferred over a network can be encrypted using conventional encryption technologies such as secure sockets layer (SSL), transport layer security (TLS), and Internet Protocol security (Ipsec).

The computing devices and systems described herein may include one or more processors, a memory, one or more storage devices, and one or more input/output (I/O) devices controllable via one or more I/O interfaces. The various components may be interconnected via at least one system bus, which may enable the transfer of data between the various modules and components of the system.

The processor(s) may be configured to process instructions for execution within the system. The processor(s) may include single-threaded processor(s), multi-threaded processor(s), or both. The processor(s) may be configured to process instructions stored in the memory or on the storage device(s). The processor(s) may include hardware-based processor(s) each including one or more cores. The processor(s) may include general purpose processor(s), special purpose processor(s), or both. The memory may store information within the system. In some implementations, the memory includes one or more computer-readable media. The memory may include any number of volatile memory units, any number of non-volatile memory units, or both volatile and non-volatile memory units. The memory may include read-only memory, random access memory, or both. In some examples, the memory may be employed as active or physical memory by one or more executing software modules.

The storage device(s) may be configured to provide (e.g., persistent) mass storage for the system. In some implementations, the storage device(s) may include one or more computer-readable media. For example, the storage device(s) may include a floppy disk device, a hard disk device, an optical disk device, or a tape device. The storage device(s) may include read-only memory, random access memory, or both. The storage device(s) may include one or more of an internal hard drive, an external hard drive, or a removable drive.

One or both of the memory or the storage device(s) may include one or more computer-readable storage media (CRSM). The CRSM may include one or more of an electronic storage medium, a magnetic storage medium, an optical storage medium, a magneto-optical storage medium, a quantum storage medium, a mechanical computer storage medium, and so forth. The CRSM may provide storage of computer-readable instructions describing data structures, processes, applications, programs, other modules, or other data for the operation of the system. In some implementations, the CRSM may include a data store that provides storage of computer-readable instructions or other information in a non-transitory format. The CRSM may be incorporated into the system or may be external with respect to the system. The CRSM may include read-only memory, random access memory, or both. One or more CRSM suitable for tangibly embodying computer program instructions and data may include any type of non-volatile memory, including but not limited to: semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. In some examples, the processor(s) and the memory may be supplemented by, or incorporated into, one or more application-specific integrated circuits (ASICs).

The system may include one or more I/O devices. The I/O device(s) may include one or more input devices such as a keyboard, a mouse, a pen, a game controller, a touch input device, an audio input device (e.g., a microphone), a gestural input device, a haptic input device, an image or video capture device (e.g., a camera), or other devices. In some examples, the I/O device(s) may also include one or more output devices such as a display, LED(s), an audio output device (e.g., a speaker), a printer, a haptic output device, and so forth. The I/O device(s) may be physically incorporated in one or more computing devices of the system, or may be external with respect to one or more computing devices of the system.

The system may include one or more I/O interfaces to enable components or modules of the system to control, interface with, or otherwise communicate with the I/O device(s). The I/O interface(s) may enable information to be transferred in or out of the system, or between components of the system, through serial communication, parallel communication, or other types of communication. For example, the I/O interface(s) may comply with a version of the RS-232 standard for serial ports, or with a version of the IEEE 1284 standard for parallel ports. As another example, the I/O interface(s) may be configured to provide a connection over Universal Serial Bus (USB) or Ethernet. In some examples, the I/O interface(s) may be configured to provide a serial connection that is compliant with a version of the IEEE 1394 standard. The I/O interface(s) may also include one or more network interfaces that enable communications between computing devices in the system, or between the system and other network-connected computing systems. The network interface(s) may include one or more network interface controllers (NICs) or other types of transceiver devices configured to send and receive communications over one or more networks, such as the network(s), using any network protocol.

Computing devices of the system may communicate with one another, or with other computing devices, using one or more networks. Such networks may include public networks such as the internet, private networks such as an institutional or personal intranet, or any combination of private and public networks. The networks may include any type of wired or wireless network, including but not limited to local area networks (LANs), wide area networks (WANs), wireless WANs (WWANs), wireless LANs (WLANs), mobile communications networks (e.g., 3G, 4G, Edge, etc.), and so forth. In some implementations, the communications between computing devices may be encrypted or otherwise secured. For example, communications may employ one or more public or private cryptographic keys, ciphers, digital certificates, or other credentials supported by a security protocol, such as any version of the Secure Sockets Layer (SSL) or the Transport Layer Security (TLS) protocol.

The system may include any number of computing devices of any type. The computing device(s) may include, but are not limited to: a personal computer, a smartphone, a tablet computer, a wearable computer, an implanted computer, a mobile gaming device, an electronic book reader, an automotive computer, a desktop computer, a laptop computer, a notebook computer, a game console, a home entertainment device, a network computer, a server computer, a mainframe computer, a distributed computing device (e.g., a cloud computing device), a microcomputer, a system on a chip (SoC), a system in a package (SiP), and so forth. Although examples herein may describe computing device(s) as physical device(s), implementations are not so limited. In some examples, a computing device may include one or more of a virtual computing environment, a hypervisor, an emulation, or a virtual machine executing on one or more physical computing devices. In some examples, two or more computing devices may include a cluster, cloud, farm, or other grouping of multiple devices that coordinate operations to provide load balancing, failover support, parallel processing capabilities, shared storage resources, shared networking capabilities, or other aspects.

Implementations and all of the functional operations described in this specification may be realized in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Implementations may be realized as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer readable medium for execution by, or to control the operation of, data processing apparatus. The computer readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, a composition of matter effecting a machine-readable propagated signal, or a combination of one or more of them. The term “computing system” encompasses all apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus may include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them. A propagated signal is an artificially generated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal that is generated to encode information for transmission to suitable receiver apparatus.

A computer program (also known as a program, software, software application, script, or code) may be written in any appropriate form of programming language, including compiled or interpreted languages, and it may be deployed in any appropriate form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program may be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code). A computer program may be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.

The processes and logic flows described in this specification may be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows may also be performed by, and apparatus may also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit).

Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any appropriate kind of digital computer. Generally, a processor may receive instructions and data from a read only memory or a random access memory or both. Elements of a computer can include a processor for performing instructions and one or more memory devices for storing instructions and data. Generally, a computer may also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks. However, a computer need not have such devices. Moreover, a computer may be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio player, a GPS receiver, to name just a few. Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto optical disks; and CD ROM and DVD-ROM disks. The processor and the memory may be supplemented by, or incorporated in, special purpose logic circuitry.

To provide for interaction with a user, implementations may be realized on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user may provide input to the computer. Other kinds of devices may be used to provide for interaction with a user as well; for example, feedback provided to the user may be any appropriate form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user may be received in any appropriate form, including acoustic, speech, or tactile input.

Implementations may be realized in a computing system that includes a back end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front end component, e.g., a client computer having a graphical user interface or a web browser through which a user may interact with an implementation, or any appropriate combination of one or more such back end, middleware, or front end components. The components of the system may be interconnected by any appropriate form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (“LAN”) and a wide area network (“WAN”), e.g., the Internet. The computing system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

While this specification contains many specifics, these should not be construed as limitations on the scope of the disclosure or of what may be claimed, but rather as descriptions of features specific to particular implementations. Certain features that are described in this specification in the context of separate implementations may also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation may also be implemented in multiple implementations separately or in any suitable sub-combination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination may in some examples be excised from the combination, and the claimed combination may be directed to a sub-combination or variation of a sub-combination.

Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the implementations described above should not be understood as requiring such separation in all implementations, and it should be understood that the described program components and systems may generally be integrated together in a single software product or packaged into multiple software products.

While various embodiments of the invention have been described, the description is intended to be exemplary, rather than limiting, and it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible that are within the scope of the invention. Accordingly, the invention is not to be restricted except in light of the attached claims and their equivalents. Also, various modifications and changes may be made within the scope of the attached claims.